fix: edge eval with var

Co-authored-by: Xu Bin <140785332+Reditiny@users.noreply.github.com>

.github/workflows/n9e.yml

@@ -5,7 +5,7 @@ on:
     tags:
       - 'v*'
 env:
-  GO_VERSION: 1.18
+  GO_VERSION: 1.23
 
 jobs:
   goreleaser:

.gitignore

@@ -9,6 +9,7 @@
 *.o
 *.a
 *.so
+*.db
 *.sw[po]
 *.tar.gz
 *.[568vq]
@@ -64,3 +65,5 @@ queries.active
 
 /n9e-*
 n9e.sql
+
+!/datasource

README.md

@@ -29,15 +29,16 @@
 
 ## 夜莺 Nightingale 是什么
 
-夜莺监控是一款开源云原生观测分析工具，采用 All-in-One 的设计理念，集数据采集、可视化、监控告警、数据分析于一体，与云原生生态紧密集成，提供开箱即用的企业级监控分析和告警能力。夜莺于 2020 年 3 月 20 日，在 github 上发布 v1 版本，已累计迭代 100 多个版本。
+夜莺监控是一款开源云原生观测分析工具，采用 All-in-One 的设计理念，集数据采集、可视化、监控告警、数据分析于一体，与云原生生态紧密集成，提供开箱即用的企业级监控分析和告警能力。夜莺于 2020 年 3 月 20 日，在 GitHub 上发布 v1 版本，已累计迭代 100 多个版本。
 
 夜莺最初由滴滴开发和开源，并于 2022 年 5 月 11 日，捐赠予中国计算机学会开源发展委员会（CCF ODC），为 CCF ODC 成立后接受捐赠的第一个开源项目。夜莺的核心研发团队，也是 Open-Falcon 项目原核心研发人员，从 2014 年（Open-Falcon 是 2014 年开源）算起来，也有 10 年了，只为把监控这个事情做好。
 
 
 ## 快速开始
-- 👉[文档中心](https://flashcat.cloud/docs/) | [下载中心](https://flashcat.cloud/download/nightingale/)
-- ❤️[报告 Bug](https://github.com/ccfos/nightingale/issues/new?assignees=&labels=&projects=&template=question.yml)
-- ℹ️为了提供更快速的访问体验，上述文档和下载站点托管于 [FlashcatCloud](https://flashcat.cloud)
+- 👉 [文档中心](https://flashcat.cloud/docs/) | [下载中心](https://flashcat.cloud/download/nightingale/)
+- ❤️ [报告 Bug](https://github.com/ccfos/nightingale/issues/new?assignees=&labels=&projects=&template=question.yml)
+- ℹ️ 为了提供更快速的访问体验，上述文档和下载站点托管于 [FlashcatCloud](https://flashcat.cloud)
+- 💡 前后端代码分离，前端代码仓库：[https://github.com/n9e/fe](https://github.com/n9e/fe)
 
 ## 功能特点
 
@@ -50,6 +51,11 @@
 
 ## 截图演示
 
+
+你可以在页面的右上角，切换语言和主题，目前我们支持英语、简体中文、繁体中文。
+
+![语言切换](https://download.flashcat.cloud/ulric/n9e-switch-i18n.png)
+
 即时查询，类似 Prometheus 内置的查询分析页面，做 ad-hoc 查询，夜莺做了一些 UI 优化，同时提供了一些内置 promql 指标，让不太了解 promql 的用户也可以快速查询。
 
 ![即时查询](https://download.flashcat.cloud/ulric/20240513103305.png)
@@ -83,15 +89,16 @@
 - 报告Bug，优先推荐提交[夜莺GitHub Issue](https://github.com/ccfos/nightingale/issues/new?assignees=&labels=kind%2Fbug&projects=&template=bug_report.yml)
 - 推荐完整浏览[夜莺文档站点](https://flashcat.cloud/docs/content/flashcat-monitor/nightingale-v7/introduction/)，了解更多信息
 - 推荐搜索关注夜莺公众号，第一时间获取社区动态：`夜莺监控Nightingale`
-- 日常问题交流推荐加入[知识星球](https://download.flashcat.cloud/ulric/20240319095409.png)，也可以加我微信 `picobyte`，备注：`夜莺加群-<公司>-<姓名>` 拉入微信群，不过研发人员主要是关注 github issue 和星球，微信群关注较少
+- 日常问题交流：
+  - QQ群：730841964
+  - [加入微信群](https://download.flashcat.cloud/ulric/20241022141621.png)，如果二维码过期了，可以联系我（我的微信：`picobyte`）拉群，备注： `夜莺互助群`
 
 ## 广受关注
 [![Stargazers over time](https://api.star-history.com/svg?repos=ccfos/nightingale&type=Date)](https://star-history.com/#ccfos/nightingale&Date)
 
-
 ## 社区共建
-- ❇️请阅读浏览[夜莺开源项目和社区治理架构草案](./doc/community-governance.md)，真诚欢迎每一位用户、开发者、公司以及组织，使用夜莺监控、积极反馈 Bug、提交功能需求、分享最佳实践，共建专业、活跃的夜莺开源社区。
-- 夜莺贡献者❤️
+- ❇️ 请阅读浏览[夜莺开源项目和社区治理架构草案](./doc/community-governance.md)，真诚欢迎每一位用户、开发者、公司以及组织，使用夜莺监控、积极反馈 Bug、提交功能需求、分享最佳实践，共建专业、活跃的夜莺开源社区。
+- ❤️ 夜莺贡献者
 <a href="https://github.com/ccfos/nightingale/graphs/contributors">
   <img src="https://contrib.rocks/image?repo=ccfos/nightingale" />
 </a>

README_en.md

@@ -1,104 +1,113 @@
 <p align="center">
   <a href="https://github.com/ccfos/nightingale">
-    <img src="doc/img/Nightingale_L_V.png" alt="nightingale - cloud native monitoring" width="240" /></a>
+    <img src="doc/img/Nightingale_L_V.png" alt="nightingale - cloud native monitoring" width="100" /></a>
+</p>
+<p align="center">
+  <b>Open-source Alert Management Expert, an Integrated Observability Platform</b>
 </p>
 
 <p align="center">
-<img alt="GitHub latest release" src="https://img.shields.io/github/v/release/ccfos/nightingale"/>
-<a href="https://n9e.github.io">
+<a href="https://flashcat.cloud/docs/">
   <img alt="Docs" src="https://img.shields.io/badge/docs-get%20started-brightgreen"/></a>
 <a href="https://hub.docker.com/u/flashcatcloud">
   <img alt="Docker pulls" src="https://img.shields.io/docker/pulls/flashcatcloud/nightingale"/></a>
-<img alt="GitHub Repo stars" src="https://img.shields.io/github/stars/ccfos/nightingale">
-<img alt="GitHub Repo issues" src="https://img.shields.io/github/issues/ccfos/nightingale">
-<img alt="GitHub Repo issues closed" src="https://img.shields.io/github/issues-closed/ccfos/nightingale">
-<img alt="GitHub forks" src="https://img.shields.io/github/forks/ccfos/nightingale">
 <a href="https://github.com/ccfos/nightingale/graphs/contributors">
   <img alt="GitHub contributors" src="https://img.shields.io/github/contributors-anon/ccfos/nightingale"/></a>
+<img alt="GitHub Repo stars" src="https://img.shields.io/github/stars/ccfos/nightingale">
+<img alt="GitHub forks" src="https://img.shields.io/github/forks/ccfos/nightingale">
+<br/><img alt="GitHub Repo issues" src="https://img.shields.io/github/issues/ccfos/nightingale">
+<img alt="GitHub Repo issues closed" src="https://img.shields.io/github/issues-closed/ccfos/nightingale">
+<img alt="GitHub latest release" src="https://img.shields.io/github/v/release/ccfos/nightingale"/>
+<img alt="License" src="https://img.shields.io/badge/license-Apache--2.0-blue"/>
 <a href="https://n9e-talk.slack.com/">
   <img alt="GitHub contributors" src="https://img.shields.io/badge/join%20slack-%23n9e-brightgreen.svg"/></a>
-<img alt="License" src="https://img.shields.io/badge/license-Apache--2.0-blue"/>
-</p>
-<p align="center">
-  An open-source cloud-native monitoring system that is <b>all-in-one</b> <br/>
-  <b>Out-of-the-box</b>, it integrates data collection, visualization, and monitoring alert <br/>
-  We recommend upgrading your <b>Prometheus + AlertManager + Grafana</b> combination to Nightingale!
 </p>
 
+
+
 [English](./README_en.md) | [中文](./README.md)
 
+## What is Nightingale
 
-## Highlighted Features
+Nightingale aims to combine the advantages of Prometheus and Grafana. It manages alert rules and visualizes metrics, logs, traces in a beautiful WebUI.
 
-- **Out-of-the-box**
-  - Supports multiple deployment methods such as **Docker, Helm Chart, and cloud services**, integrates data collection, monitoring, and alerting into one system, and comes with various monitoring dashboards, quick views, and alert rule templates. **It greatly reduces the construction cost, learning cost, and usage cost of cloud-native monitoring systems**.
-- **Professional Alerting**
-  - Provides visual alert configuration and management, supports various alert rules, offers the ability to configure silence and subscription rules, supports multiple alert delivery channels, and has features such as alert self-healing and event management.
-- **Cloud-Native**
-  - Quickly builds an enterprise-level cloud-native monitoring system through a turnkey approach, supports multiple collectors such as [Categraf](https://github.com/flashcatcloud/categraf), Telegraf, and Grafana-agent, supports multiple data sources such as Prometheus, VictoriaMetrics, M3DB, ElasticSearch, and Jaeger, and is compatible with importing Grafana dashboards. **It seamlessly integrates with the cloud-native ecosystem**.
-- **High Performance and High Availability**
-  - Due to the multi-data-source management engine of Nightingale and its excellent architecture design, and utilizing a high-performance time-series database, it can handle data collection, storage, and alert analysis scenarios with billions of time-series data, saving a lot of costs.
-  - Nightingale components can be horizontally scaled with no single point of failure. It has been deployed in thousands of enterprises and tested in harsh production practices. Many leading Internet companies have used Nightingale for cluster machines with hundreds of nodes, processing billions of time-series data.
-- **Flexible Extension and Centralized Management**
-  - Nightingale can be deployed on a 1-core 1G cloud host, deployed in a cluster of hundreds of machines, or run in Kubernetes. Time-series databases, alert engines, and other components can also be decentralized to various data centers and regions, balancing edge deployment with centralized management. **It solves the problem of data fragmentation and lack of unified views**.
+Originally developed and open-sourced by Didi, Nightingale was donated to the China Computer Federation Open Source Development Committee (CCF ODC) on May 11, 2022, becoming the first open-source project accepted by the CCF ODC after its establishment. 
 
 
-#### If you are using Prometheus and have one or more of the following requirement scenarios, it is recommended that you upgrade to Nightingale:
+## Quick Start
 
-- Multiple systems such as Prometheus, Alertmanager, Grafana, etc. are fragmented and lack a unified view and cannot be used out of the box;
-- The way to manage Prometheus and Alertmanager by modifying configuration files has a big learning curve and is difficult to collaborate;
-- Too much data to scale-up your Prometheus cluster;
-- Multiple Prometheus clusters running in production environments, which faced high management and usage costs;
+- 👉 [Documentation](https://flashcat.cloud/docs/) | [Download](https://flashcat.cloud/download/nightingale/)
+- ❤️ [Report a Bug](https://github.com/ccfos/nightingale/issues/new?assignees=&labels=&projects=&template=question.yml)
+- ℹ️ For faster access, the above documentation and download sites are hosted on [FlashcatCloud](https://flashcat.cloud).
 
-#### If you are using Zabbix and have the following scenarios, it is recommended that you upgrade to Nightingale:
+## Features
 
-- Monitoring too much data and wanting a better scalable solution;
-- A high learning curve and a desire for better efficiency of collaborative use in a multi-person, multi-team model;
-- Microservice and cloud-native architectures with variable monitoring data lifecycles and high monitoring data dimension bases, which are not easily adaptable to the Zabbix data model;
+- **Integration with Multiple Time-Series Databases:** Supports integration with various time-series databases such as Prometheus, VictoriaMetrics, Thanos, Mimir, M3DB, and TDengine, enabling unified alert management.
+- **Advanced Alerting Capabilities:** Comes with built-in support for multiple alerting rules, extensible to common notification channels. It also supports alert suppression, silencing, subscription, self-healing, and alert event management.
+- **High-Performance Visualization Engine:** Offers various chart styles with numerous built-in dashboard templates and the ability to import Grafana templates. Ready to use with a business-friendly open-source license.
+- **Support for Common Collectors:** Compatible with [Categraf](https://flashcat.cloud/product/categraf), Telegraf, Grafana-agent, Datadog-agent, and various exporters as collectors—there's no data that can't be monitored.
+- **Seamless Integration with [Flashduty](https://flashcat.cloud/product/flashcat-duty/):** Enables alert aggregation, acknowledgment, escalation, scheduling, and IM integration, ensuring no alerts are missed, reducing unnecessary interruptions, and enhancing efficient collaboration.
 
 
-#### If you are using [open-falcon](https://github.com/open-falcon/falcon-plus), we recommend you to upgrade to Nightingale：
-- For more information about open-falcon and Nightingale, please refer to read [Ten features and trends of cloud-native monitoring](https://mp.weixin.qq.com/s?__biz=MzkzNjI5OTM5Nw==&mid=2247483738&idx=1&sn=e8bdbb974a2cd003c1abcc2b5405dd18&chksm=c2a19fb0f5d616a63185cd79277a79a6b80118ef2185890d0683d2bb20451bd9303c78d083c5#rd)。
-
-## Getting Started
-
-[https://n9e.github.io/](https://n9e.github.io/)
-
 ## Screenshots
 
-https://user-images.githubusercontent.com/792850/216888712-2565fcea-9df5-47bd-a49e-d60af9bd76e8.mp4
+You can switch languages and themes in the top right corner. We now support English, Simplified Chinese, and Traditional Chinese. 
+
+![18n switch](https://download.flashcat.cloud/ulric/n9e-switch-i18n.png)
+
+### Instant Query
+
+Similar to the built-in query analysis page in Prometheus, Nightingale offers an ad-hoc query feature with UI enhancements. It also provides built-in PromQL metrics, allowing users unfamiliar with PromQL to quickly perform queries.
+
+![Instant Query](https://download.flashcat.cloud/ulric/20240513103305.png)
+
+### Metric View
+
+Alternatively, you can use the Metric View to access data. With this feature, Instant Query becomes less necessary, as it caters more to advanced users. Regular users can easily perform queries using the Metric View.
+
+![Metric View](https://download.flashcat.cloud/ulric/20240513103530.png)
+
+### Built-in Dashboards
+
+Nightingale includes commonly used dashboards that can be imported and used directly. You can also import Grafana dashboards, although compatibility is limited to basic Grafana charts. If you’re accustomed to Grafana, it’s recommended to continue using it for visualization, with Nightingale serving as an alerting engine.
+
+![Built-in Dashboards](https://download.flashcat.cloud/ulric/20240513103628.png)
+
+### Built-in Alert Rules
+
+In addition to the built-in dashboards, Nightingale also comes with numerous alert rules that are ready to use out of the box.
+
+![Built-in Alert Rules](https://download.flashcat.cloud/ulric/20240513103825.png)
+
+
 
 ## Architecture
 
-<img src="doc/img/arch-product.png" width="600">
+In most community scenarios, Nightingale is primarily used as an alert engine, integrating with multiple time-series databases to unify alert rule management. Grafana remains the preferred tool for visualization. As an alert engine, the product architecture of Nightingale is as follows:
 
-Nightingale monitoring can receive monitoring data reported by various collectors (such as [Categraf](https://github.com/flashcatcloud/categraf) , telegraf, grafana-agent, Prometheus, etc.) and write them to various popular time-series databases (such as Prometheus, M3DB, VictoriaMetrics, Thanos, TDEngine, etc.). It provides configuration capabilities for alert rules, silence rules, and subscription rules, as well as the ability to view monitoring data. It also provides automatic alarm self-healing mechanisms (such as automatically calling back to a webhook address or executing a script after an alarm is triggered), and the ability to store and manage historical alarm events and view them in groups.
+![Product Architecture](https://download.flashcat.cloud/ulric/20240221152601.png)
 
-If the performance of a standalone time-series database (such as Prometheus) has bottlenecks or poor disaster recovery, we recommend using [VictoriaMetrics](https://github.com/VictoriaMetrics/VictoriaMetrics). The VictoriaMetrics architecture is relatively simple, has excellent performance, and is easy to deploy and maintain. The architecture diagram is as shown above. For more detailed documentation on VictoriaMetrics, please refer to its [official website](https://victoriametrics.com/).
+For certain edge data centers with poor network connectivity to the central Nightingale server, we offer a distributed deployment mode for the alert engine. In this mode, even if the network is disconnected, the alerting functionality remains unaffected.
 
-**We welcome you to participate in the Nightingale open-source project and community in various ways, including but not limited to**：
-- Adding and improving documentation => [n9e.github.io](https://n9e.github.io/)
-- Sharing your best practices and experience in using Nightingale monitoring => [Article sharing]((https://n9e.github.io/docs/prologue/share/))
-- Submitting product suggestions => [github issue](https://github.com/ccfos/nightingale/issues/new?assignees=&labels=kind%2Ffeature&template=enhancement.md)
-- Submitting code to make Nightingale monitoring faster, more stable, and easier to use => [github pull request](https://github.com/didi/nightingale/pulls)
+![Edge Deployment Mode](https://download.flashcat.cloud/ulric/20240222102119.png)
 
 
-**Respecting, recognizing, and recording the work of every contributor** is the first guiding principle of the Nightingale open-source community. We advocate effective questioning, which not only respects the developer's time but also contributes to the accumulation of knowledge in the entire community
-- Before asking a question, please first refer to the [FAQ](https://www.gitlink.org.cn/ccfos/nightingale/wiki/faq) 
-- We use [GitHub Discussions](https://github.com/ccfos/nightingale/discussions) as the communication forum. You can search and ask questions here.
-- We also recommend that you join ours [Slack channel](https://n9e-talk.slack.com/) to exchange experiences with other Nightingale users.
+## Communication Channels
 
-
-## Who is using Nightingale
-You can register your usage and share your experience by posting on **[Who is Using Nightingale](https://github.com/ccfos/nightingale/issues/897)**.
+- **Report Bugs:** It is highly recommended to submit issues via the [Nightingale GitHub Issue tracker](https://github.com/ccfos/nightingale/issues/new?assignees=&labels=kind%2Fbug&projects=&template=bug_report.yml).
+- **Documentation:** For more information, we recommend thoroughly browsing the [Nightingale Documentation Site](https://flashcat.cloud/docs/content/flashcat-monitor/nightingale-v7/introduction/).
 
 ## Stargazers over time
-[![Stargazers over time](https://starchart.cc/ccfos/nightingale.svg)](https://starchart.cc/ccfos/nightingale)
 
-## Contributors
+[![Stargazers over time](https://api.star-history.com/svg?repos=ccfos/nightingale&type=Date)](https://star-history.com/#ccfos/nightingale&Date)
+
+## Community Co-Building
+
+- ❇️ Please read the [Nightingale Open Source Project and Community Governance Draft](./doc/community-governance.md). We sincerely welcome every user, developer, company, and organization to use Nightingale, actively report bugs, submit feature requests, share best practices, and help build a professional and active open-source community.
+-  ❤️ Nightingale Contributors
 <a href="https://github.com/ccfos/nightingale/graphs/contributors">
   <img src="https://contrib.rocks/image?repo=ccfos/nightingale" />
 </a>
 
 ## License
-[Apache License V2.0](https://github.com/didi/nightingale/blob/main/LICENSE)
+- [Apache License V2.0](https://github.com/didi/nightingale/blob/main/LICENSE)

alert/aconf/conf.go

@@ -32,6 +32,7 @@ type Alerting struct {
 	Timeout           int64
 	TemplatesDir      string
 	NotifyConcurrency int
+	WebhookBatchSend  bool
 }
 
 type CallPlugin struct {
@@ -59,10 +60,6 @@ func (a *Alert) PreCheck(configDir string) {
 		a.Heartbeat.Interval = 1000
 	}
 
-	if a.Heartbeat.EngineName == "" {
-		a.Heartbeat.EngineName = "default"
-	}
-
 	if a.EngineDelay == 0 {
 		a.EngineDelay = 30
 	}

alert/alert.go

@@ -4,6 +4,8 @@ import (
 	"context"
 	"fmt"
 
+	"github.com/ccfos/nightingale/v6/dscache"
+
 	"github.com/ccfos/nightingale/v6/alert/aconf"
 	"github.com/ccfos/nightingale/v6/alert/astats"
 	"github.com/ccfos/nightingale/v6/alert/dispatch"
@@ -21,12 +23,11 @@ import (
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/pkg/httpx"
 	"github.com/ccfos/nightingale/v6/pkg/logx"
+	"github.com/ccfos/nightingale/v6/pkg/macros"
 	"github.com/ccfos/nightingale/v6/prom"
 	"github.com/ccfos/nightingale/v6/pushgw/pconf"
 	"github.com/ccfos/nightingale/v6/pushgw/writer"
 	"github.com/ccfos/nightingale/v6/storage"
-	"github.com/ccfos/nightingale/v6/tdengine"
-
 	"github.com/flashcatcloud/ibex/src/cmd/ibex"
 )
 
@@ -62,15 +63,19 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 	userCache := memsto.NewUserCache(ctx, syncStats)
 	userGroupCache := memsto.NewUserGroupCache(ctx, syncStats)
 	taskTplsCache := memsto.NewTaskTplCache(ctx)
+	configCvalCache := memsto.NewCvalCache(ctx, syncStats)
 
 	promClients := prom.NewPromClient(ctx)
-	tdengineClients := tdengine.NewTdengineClient(ctx, config.Alert.Heartbeat)
+	dispatch.InitRegisterQueryFunc(promClients)
 
 	externalProcessors := process.NewExternalProcessors()
 
-	Start(config.Alert, config.Pushgw, syncStats, alertStats, externalProcessors, targetCache, busiGroupCache, alertMuteCache, alertRuleCache, notifyConfigCache, taskTplsCache, dsCache, ctx, promClients, tdengineClients, userCache, userGroupCache)
+	macros.RegisterMacro(macros.MacroInVain)
+	dscache.Init(ctx, false)
+	Start(config.Alert, config.Pushgw, syncStats, alertStats, externalProcessors, targetCache, busiGroupCache, alertMuteCache, alertRuleCache, notifyConfigCache, taskTplsCache, dsCache, ctx, promClients, userCache, userGroupCache)
 
-	r := httpx.GinEngine(config.Global.RunMode, config.HTTP)
+	r := httpx.GinEngine(config.Global.RunMode, config.HTTP,
+		configCvalCache.PrintBodyPaths, configCvalCache.PrintAccessLog)
 	rt := router.New(config.HTTP, config.Alert, alertMuteCache, targetCache, busiGroupCache, alertStats, ctx, externalProcessors)
 
 	if config.Ibex.Enable {
@@ -90,7 +95,7 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 
 func Start(alertc aconf.Alert, pushgwc pconf.Pushgw, syncStats *memsto.Stats, alertStats *astats.Stats, externalProcessors *process.ExternalProcessorsType, targetCache *memsto.TargetCacheType, busiGroupCache *memsto.BusiGroupCacheType,
 	alertMuteCache *memsto.AlertMuteCacheType, alertRuleCache *memsto.AlertRuleCacheType, notifyConfigCache *memsto.NotifyConfigCacheType, taskTplsCache *memsto.TaskTplCache, datasourceCache *memsto.DatasourceCacheType, ctx *ctx.Context,
-	promClients *prom.PromClientMap, tdendgineClients *tdengine.TdengineClientMap, userCache *memsto.UserCacheType, userGroupCache *memsto.UserGroupCacheType) {
+	promClients *prom.PromClientMap, userCache *memsto.UserCacheType, userGroupCache *memsto.UserGroupCacheType) {
 	alertSubscribeCache := memsto.NewAlertSubscribeCache(ctx, syncStats)
 	recordingRuleCache := memsto.NewRecordingRuleCache(ctx, syncStats)
 	targetsOfAlertRulesCache := memsto.NewTargetOfAlertRuleCache(ctx, alertc.Heartbeat.EngineName, syncStats)
@@ -100,10 +105,10 @@ func Start(alertc aconf.Alert, pushgwc pconf.Pushgw, syncStats *memsto.Stats, al
 	naming := naming.NewNaming(ctx, alertc.Heartbeat, alertStats)
 
 	writers := writer.NewWriters(pushgwc)
-	record.NewScheduler(alertc, recordingRuleCache, promClients, writers, alertStats)
+	record.NewScheduler(alertc, recordingRuleCache, promClients, writers, alertStats, datasourceCache)
 
 	eval.NewScheduler(alertc, externalProcessors, alertRuleCache, targetCache, targetsOfAlertRulesCache,
-		busiGroupCache, alertMuteCache, datasourceCache, promClients, tdendgineClients, naming, ctx, alertStats)
+		busiGroupCache, alertMuteCache, datasourceCache, promClients, naming, ctx, alertStats)
 
 	dp := dispatch.NewDispatch(alertRuleCache, userCache, userGroupCache, alertSubscribeCache, targetCache, notifyConfigCache, taskTplsCache, alertc.Alerting, ctx, alertStats)
 	consumer := dispatch.NewConsumer(alertc.Alerting, ctx, dp, promClients)
@@ -112,5 +117,5 @@ func Start(alertc aconf.Alert, pushgwc pconf.Pushgw, syncStats *memsto.Stats, al
 	go consumer.LoopConsume()
 
 	go queue.ReportQueueSize(alertStats)
-	go sender.InitEmailSender(notifyConfigCache)
+	go sender.InitEmailSender(ctx, notifyConfigCache)
 }

alert/astats/stats.go

@@ -38,7 +38,7 @@ func NewSyncStats() *Stats {
 		Subsystem: subsystem,
 		Name:      "rule_eval_error_total",
 		Help:      "Number of rule eval error.",
-	}, []string{"datasource", "stage"})
+	}, []string{"datasource", "stage", "busi_group", "rule_id"})
 
 	CounterQueryDataErrorTotal := prometheus.NewCounterVec(prometheus.CounterOpts{
 		Namespace: namespace,

alert/common/key.go

@@ -2,6 +2,7 @@ package common
 
 import (
 	"fmt"
+	"strings"
 
 	"github.com/ccfos/nightingale/v6/models"
 )
@@ -34,9 +35,9 @@ func MatchGroupsName(groupName string, groupFilter []models.TagFilter) bool {
 func matchTag(value string, filter models.TagFilter) bool {
 	switch filter.Func {
 	case "==":
-		return filter.Value == value
+		return strings.TrimSpace(filter.Value) == strings.TrimSpace(value)
 	case "!=":
-		return filter.Value != value
+		return strings.TrimSpace(filter.Value) != strings.TrimSpace(value)
 	case "in":
 		_, has := filter.Vset[value]
 		return has

alert/dispatch/consume.go

@@ -1,6 +1,7 @@
 package dispatch
 
 import (
+	"context"
 	"encoding/json"
 	"fmt"
 	"strings"
@@ -13,8 +14,10 @@ import (
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/pkg/poster"
 	promsdk "github.com/ccfos/nightingale/v6/pkg/prom"
+	"github.com/ccfos/nightingale/v6/pkg/tplx"
 	"github.com/ccfos/nightingale/v6/prom"
 
+	"github.com/prometheus/common/model"
 	"github.com/toolkits/pkg/concurrent/semaphore"
 	"github.com/toolkits/pkg/logger"
 )
@@ -27,6 +30,18 @@ type Consumer struct {
 	promClients *prom.PromClientMap
 }
 
+func InitRegisterQueryFunc(promClients *prom.PromClientMap) {
+	tplx.RegisterQueryFunc(func(datasourceID int64, promql string) model.Value {
+		if promClients.IsNil(datasourceID) {
+			return nil
+		}
+
+		readerClient := promClients.GetCli(datasourceID)
+		value, _, _ := readerClient.Query(context.Background(), promql, time.Now())
+		return value
+	})
+}
+
 // 创建一个 Consumer 实例
 func NewConsumer(alerting aconf.Alerting, ctx *ctx.Context, dispatch *Dispatch, promClients *prom.PromClientMap) *Consumer {
 	return &Consumer{
@@ -113,7 +128,7 @@ func (e *Consumer) persist(event *models.AlertCurEvent) {
 		event.Id, err = poster.PostByUrlsWithResp[int64](e.ctx, "/v1/n9e/event-persist", event)
 		if err != nil {
 			logger.Errorf("event:%+v persist err:%v", event, err)
-			e.dispatch.Astats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", event.DatasourceId), "persist_event").Inc()
+			e.dispatch.Astats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", event.DatasourceId), "persist_event", event.GroupName, fmt.Sprintf("%v", event.RuleId)).Inc()
 		}
 		return
 	}
@@ -121,7 +136,7 @@ func (e *Consumer) persist(event *models.AlertCurEvent) {
 	err := models.EventPersist(e.ctx, event)
 	if err != nil {
 		logger.Errorf("event%+v persist err:%v", event, err)
-		e.dispatch.Astats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", event.DatasourceId), "persist_event").Inc()
+		e.dispatch.Astats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", event.DatasourceId), "persist_event", event.GroupName, fmt.Sprintf("%v", event.RuleId)).Inc()
 	}
 }
 
@@ -169,7 +184,7 @@ func (e *Consumer) queryRecoveryVal(event *models.AlertCurEvent) {
 		logger.Errorf("rule_eval:%s promql:%s, warnings:%v", getKey(event), promql, warnings)
 	}
 
-	anomalyPoints := common.ConvertAnomalyPoints(value)
+	anomalyPoints := models.ConvertAnomalyPoints(value)
 	if len(anomalyPoints) == 0 {
 		logger.Warningf("rule_eval:%s promql:%s, result is empty", getKey(event), promql)
 		event.AnnotationsJSON["recovery_promql_error"] = fmt.Sprintf("promql:%s error:%s", promql, "result is empty")

alert/dispatch/dispatch.go

@@ -139,6 +139,12 @@ func (e *Dispatch) HandleEventNotify(event *models.AlertCurEvent, isSubscribe bo
 	if rule == nil {
 		return
 	}
+
+	if e.blockEventNotify(rule, event) {
+		logger.Infof("block event notify: rule_id:%d event:%+v", rule.Id, event)
+		return
+	}
+
 	fillUsers(event, e.userCache, e.userGroupCache)
 
 	var (
@@ -167,7 +173,7 @@ func (e *Dispatch) HandleEventNotify(event *models.AlertCurEvent, isSubscribe bo
 	}
 
 	// 处理事件发送,这里用一个goroutine处理一个event的所有发送事件
-	go e.Send(rule, event, notifyTarget)
+	go e.Send(rule, event, notifyTarget, isSubscribe)
 
 	// 如果是不是订阅规则出现的event, 则需要处理订阅规则的event
 	if !isSubscribe {
@@ -175,6 +181,25 @@ func (e *Dispatch) HandleEventNotify(event *models.AlertCurEvent, isSubscribe bo
 	}
 }
 
+func (e *Dispatch) blockEventNotify(rule *models.AlertRule, event *models.AlertCurEvent) bool {
+	ruleType := rule.GetRuleType()
+
+	// 若为机器则先看机器是否删除
+	if ruleType == models.HOST {
+		host, ok := e.targetCache.Get(event.TagsMap["ident"])
+		if !ok || host == nil {
+			return true
+		}
+	}
+
+	// 恢复通知，检测规则配置是否改变
+	// if event.IsRecovered && event.RuleHash != rule.Hash() {
+	// 	return true
+	// }
+
+	return false
+}
+
 func (e *Dispatch) handleSubs(event *models.AlertCurEvent) {
 	// handle alert subscribes
 	subscribes := make([]*models.AlertSubscribe, 0)
@@ -238,11 +263,12 @@ func (e *Dispatch) handleSub(sub *models.AlertSubscribe, event models.AlertCurEv
 	e.HandleEventNotify(&event, true)
 }
 
-func (e *Dispatch) Send(rule *models.AlertRule, event *models.AlertCurEvent, notifyTarget *NotifyTarget) {
+func (e *Dispatch) Send(rule *models.AlertRule, event *models.AlertCurEvent, notifyTarget *NotifyTarget, isSubscribe bool) {
 	needSend := e.BeforeSenderHook(event)
 	if needSend {
 		for channel, uids := range notifyTarget.ToChannelUserMap() {
-			msgCtx := sender.BuildMessageContext(rule, []*models.AlertCurEvent{event}, uids, e.userCache, e.Astats)
+			msgCtx := sender.BuildMessageContext(e.ctx, rule, []*models.AlertCurEvent{event},
+				uids, e.userCache, e.Astats)
 			e.RwLock.RLock()
 			s := e.Senders[channel]
 			e.RwLock.RUnlock()
@@ -265,22 +291,40 @@ func (e *Dispatch) Send(rule *models.AlertRule, event *models.AlertCurEvent, not
 	e.SendCallbacks(rule, notifyTarget, event)
 
 	// handle global webhooks
-	sender.SendWebhooks(notifyTarget.ToWebhookList(), event, e.Astats)
+	if !event.OverrideGlobalWebhook() {
+		if e.alerting.WebhookBatchSend {
+			sender.BatchSendWebhooks(e.ctx, notifyTarget.ToWebhookMap(), event, e.Astats)
+		} else {
+			sender.SingleSendWebhooks(e.ctx, notifyTarget.ToWebhookMap(), event, e.Astats)
+		}
+	}
 
 	// handle plugin call
-	go sender.MayPluginNotify(e.genNoticeBytes(event), e.notifyConfigCache.GetNotifyScript(), e.Astats)
+	go sender.MayPluginNotify(e.ctx, e.genNoticeBytes(event), e.notifyConfigCache.
+		GetNotifyScript(), e.Astats, event)
+
+	if !isSubscribe {
+		// handle ibex callbacks
+		e.HandleIbex(rule, event)
+	}
 }
 
 func (e *Dispatch) SendCallbacks(rule *models.AlertRule, notifyTarget *NotifyTarget, event *models.AlertCurEvent) {
-
 	uids := notifyTarget.ToUidList()
 	urls := notifyTarget.ToCallbackList()
+	whMap := notifyTarget.ToWebhookMap()
+	ogw := event.OverrideGlobalWebhook()
 	for _, urlStr := range urls {
 		if len(urlStr) == 0 {
 			continue
 		}
 
-		cbCtx := sender.BuildCallBackContext(e.ctx, urlStr, rule, []*models.AlertCurEvent{event}, uids, e.userCache, e.Astats)
+		cbCtx := sender.BuildCallBackContext(e.ctx, urlStr, rule, []*models.AlertCurEvent{event}, uids, e.userCache, e.alerting.WebhookBatchSend, e.Astats)
+
+		if wh, ok := whMap[cbCtx.CallBackURL]; !ogw && ok && wh.Enable {
+			logger.Debugf("SendCallbacks: webhook[%s] is in global conf.", cbCtx.CallBackURL)
+			continue
+		}
 
 		if strings.HasPrefix(urlStr, "${ibex}") {
 			e.CallBacks[models.IbexDomain].CallBack(cbCtx)
@@ -318,6 +362,30 @@ func (e *Dispatch) SendCallbacks(rule *models.AlertRule, notifyTarget *NotifyTar
 	}
 }
 
+func (e *Dispatch) HandleIbex(rule *models.AlertRule, event *models.AlertCurEvent) {
+	// 解析 RuleConfig 字段
+	var ruleConfig struct {
+		TaskTpls []*models.Tpl `json:"task_tpls"`
+	}
+	json.Unmarshal([]byte(rule.RuleConfig), &ruleConfig)
+
+	for _, t := range ruleConfig.TaskTpls {
+		if t.TplId == 0 {
+			continue
+		}
+
+		if len(t.Host) == 0 {
+			sender.CallIbex(e.ctx, t.TplId, event.TargetIdent,
+				e.taskTplsCache, e.targetCache, e.userCache, event)
+			continue
+		}
+		for _, host := range t.Host {
+			sender.CallIbex(e.ctx, t.TplId, host,
+				e.taskTplsCache, e.targetCache, e.userCache, event)
+		}
+	}
+}
+
 type Notice struct {
 	Event *models.AlertCurEvent `json:"event"`
 	Tpls  map[string]string     `json:"tpls"`

alert/dispatch/notify_target.go

@@ -76,16 +76,12 @@ func (s *NotifyTarget) ToCallbackList() []string {
 	return callbacks
 }
 
-func (s *NotifyTarget) ToWebhookList() []*models.Webhook {
-	webhooks := make([]*models.Webhook, 0, len(s.webhooks))
-	for _, wh := range s.webhooks {
-		webhooks = append(webhooks, wh)
-	}
-	return webhooks
+func (s *NotifyTarget) ToWebhookMap() map[string]*models.Webhook {
+	return s.webhooks
 }
 
 func (s *NotifyTarget) ToUidList() []int64 {
-	uids := make([]int64, len(s.userMap))
+	uids := make([]int64, 0, len(s.userMap))
 	for uid, _ := range s.userMap {
 		uids = append(uids, uid)
 	}

alert/eval/alert_rule.go

@@ -13,8 +13,6 @@ import (
 	"github.com/ccfos/nightingale/v6/memsto"
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/prom"
-	"github.com/ccfos/nightingale/v6/tdengine"
-
 	"github.com/toolkits/pkg/logger"
 )
 
@@ -33,8 +31,7 @@ type Scheduler struct {
 	alertMuteCache          *memsto.AlertMuteCacheType
 	datasourceCache         *memsto.DatasourceCacheType
 
-	promClients     *prom.PromClientMap
-	tdengineClients *tdengine.TdengineClientMap
+	promClients *prom.PromClientMap
 
 	naming *naming.Naming
 
@@ -45,7 +42,7 @@ type Scheduler struct {
 func NewScheduler(aconf aconf.Alert, externalProcessors *process.ExternalProcessorsType, arc *memsto.AlertRuleCacheType,
 	targetCache *memsto.TargetCacheType, toarc *memsto.TargetsOfAlertRuleCacheType,
 	busiGroupCache *memsto.BusiGroupCacheType, alertMuteCache *memsto.AlertMuteCacheType, datasourceCache *memsto.DatasourceCacheType,
-	promClients *prom.PromClientMap, tdengineClients *tdengine.TdengineClientMap, naming *naming.Naming, ctx *ctx.Context, stats *astats.Stats) *Scheduler {
+	promClients *prom.PromClientMap, naming *naming.Naming, ctx *ctx.Context, stats *astats.Stats) *Scheduler {
 	scheduler := &Scheduler{
 		aconf:      aconf,
 		alertRules: make(map[string]*AlertRuleWorker),
@@ -59,9 +56,8 @@ func NewScheduler(aconf aconf.Alert, externalProcessors *process.ExternalProcess
 		alertMuteCache:          alertMuteCache,
 		datasourceCache:         datasourceCache,
 
-		promClients:     promClients,
-		tdengineClients: tdengineClients,
-		naming:          naming,
+		promClients: promClients,
+		naming:      naming,
 
 		ctx:   ctx,
 		stats: stats,
@@ -95,9 +91,8 @@ func (s *Scheduler) syncAlertRules() {
 		}
 
 		ruleType := rule.GetRuleType()
-		if rule.IsPrometheusRule() || rule.IsLokiRule() || rule.IsTdengineRule() {
-			datasourceIds := s.promClients.Hit(rule.DatasourceIdsJson)
-			datasourceIds = append(datasourceIds, s.tdengineClients.Hit(rule.DatasourceIdsJson)...)
+		if rule.IsPrometheusRule() || rule.IsLokiRule() || rule.IsTdengineRule() || rule.IsClickHouseRule() || rule.IsElasticSearch() {
+			datasourceIds := s.datasourceCache.GetIDsByDsCateAndQueries(rule.Cate, rule.DatasourceQueries)
 			for _, dsId := range datasourceIds {
 				if !naming.DatasourceHashRing.IsHit(strconv.FormatInt(dsId, 10), fmt.Sprintf("%d", rule.Id), s.aconf.Heartbeat.Endpoint) {
 					continue
@@ -119,7 +114,7 @@ func (s *Scheduler) syncAlertRules() {
 				}
 				processor := process.NewProcessor(s.aconf.Heartbeat.EngineName, rule, dsId, s.alertRuleCache, s.targetCache, s.targetsOfAlertRuleCache, s.busiGroupCache, s.alertMuteCache, s.datasourceCache, s.ctx, s.stats)
 
-				alertRule := NewAlertRuleWorker(rule, dsId, processor, s.promClients, s.tdengineClients, s.ctx)
+				alertRule := NewAlertRuleWorker(rule, dsId, processor, s.promClients, s.ctx)
 				alertRuleWorkers[alertRule.Hash()] = alertRule
 			}
 		} else if rule.IsHostRule() {
@@ -128,12 +123,13 @@ func (s *Scheduler) syncAlertRules() {
 				continue
 			}
 			processor := process.NewProcessor(s.aconf.Heartbeat.EngineName, rule, 0, s.alertRuleCache, s.targetCache, s.targetsOfAlertRuleCache, s.busiGroupCache, s.alertMuteCache, s.datasourceCache, s.ctx, s.stats)
-			alertRule := NewAlertRuleWorker(rule, 0, processor, s.promClients, s.tdengineClients, s.ctx)
+			alertRule := NewAlertRuleWorker(rule, 0, processor, s.promClients, s.ctx)
 			alertRuleWorkers[alertRule.Hash()] = alertRule
 		} else {
 			// 如果 rule 不是通过 prometheus engine 来告警的，则创建为 externalRule
 			// if rule is not processed by prometheus engine, create it as externalRule
-			for _, dsId := range rule.DatasourceIdsJson {
+			dsIds := s.datasourceCache.GetIDsByDsCateAndQueries(rule.Cate, rule.DatasourceQueries)
+			for _, dsId := range dsIds {
 				ds := s.datasourceCache.GetById(dsId)
 				if ds == nil {
 					logger.Debugf("datasource %d not found", dsId)
@@ -153,6 +149,7 @@ func (s *Scheduler) syncAlertRules() {
 	for hash, rule := range alertRuleWorkers {
 		if _, has := s.alertRules[hash]; !has {
 			rule.Prepare()
+			time.Sleep(time.Duration(20) * time.Millisecond)
 			rule.Start()
 			s.alertRules[hash] = rule
 		}

alert/eval/eval_test.go

@@ -0,0 +1,458 @@
+package eval
+
+import (
+	"reflect"
+	"testing"
+
+	"golang.org/x/exp/slices"
+)
+
+var (
+	reHashTagIndex1 = map[uint64][][]uint64{
+		1: {
+			{1, 2}, {3, 4},
+		},
+		2: {
+			{5, 6}, {7, 8},
+		},
+	}
+	reHashTagIndex2 = map[uint64][][]uint64{
+		1: {
+			{9, 10}, {11, 12},
+		},
+		3: {
+			{13, 14}, {15, 16},
+		},
+	}
+	seriesTagIndex1 = map[uint64][]uint64{
+		1: {1, 2, 3, 4},
+		2: {5, 6, 7, 8},
+	}
+	seriesTagIndex2 = map[uint64][]uint64{
+		1: {9, 10, 11, 12},
+		3: {13, 14, 15, 16},
+	}
+)
+
+func Test_originalJoin(t *testing.T) {
+	type args struct {
+		seriesTagIndex1 map[uint64][]uint64
+		seriesTagIndex2 map[uint64][]uint64
+	}
+	tests := []struct {
+		name string
+		args args
+		want map[uint64][]uint64
+	}{
+		{
+			name: "original join",
+			args: args{
+				seriesTagIndex1: map[uint64][]uint64{
+					1: {1, 2, 3, 4},
+					2: {5, 6, 7, 8},
+				},
+				seriesTagIndex2: map[uint64][]uint64{
+					1: {9, 10, 11, 12},
+					3: {13, 14, 15, 16},
+				},
+			},
+			want: map[uint64][]uint64{
+				1: {1, 2, 3, 4, 9, 10, 11, 12},
+				2: {5, 6, 7, 8},
+				3: {13, 14, 15, 16},
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := originalJoin(tt.args.seriesTagIndex1, tt.args.seriesTagIndex2); !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("originalJoin() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_exclude(t *testing.T) {
+	type args struct {
+		reHashTagIndex1 map[uint64][][]uint64
+		reHashTagIndex2 map[uint64][][]uint64
+	}
+	tests := []struct {
+		name string
+		args args
+		want map[uint64][]uint64
+	}{
+		{
+			name: "left exclude",
+			args: args{
+				reHashTagIndex1: reHashTagIndex1,
+				reHashTagIndex2: reHashTagIndex2,
+			},
+			want: map[uint64][]uint64{
+				0: {5, 6},
+				1: {7, 8},
+			},
+		},
+		{
+			name: "right exclude",
+			args: args{
+				reHashTagIndex1: reHashTagIndex2,
+				reHashTagIndex2: reHashTagIndex1,
+			},
+			want: map[uint64][]uint64{
+				3: {13, 14},
+				4: {15, 16},
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := exclude(tt.args.reHashTagIndex1, tt.args.reHashTagIndex2); !allValueDeepEqual(flatten(got), tt.want) {
+				t.Errorf("exclude() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_noneJoin(t *testing.T) {
+	type args struct {
+		seriesTagIndex1 map[uint64][]uint64
+		seriesTagIndex2 map[uint64][]uint64
+	}
+	tests := []struct {
+		name string
+		args args
+		want map[uint64][]uint64
+	}{
+		{
+			name: "none join, direct splicing",
+			args: args{
+				seriesTagIndex1: seriesTagIndex1,
+				seriesTagIndex2: seriesTagIndex2,
+			},
+			want: map[uint64][]uint64{
+				0: {1, 2, 3, 4},
+				1: {5, 6, 7, 8},
+				2: {9, 10, 11, 12},
+				3: {13, 14, 15, 16},
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := noneJoin(tt.args.seriesTagIndex1, tt.args.seriesTagIndex2); !allValueDeepEqual(got, tt.want) {
+				t.Errorf("noneJoin() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_cartesianJoin(t *testing.T) {
+	type args struct {
+		seriesTagIndex1 map[uint64][]uint64
+		seriesTagIndex2 map[uint64][]uint64
+	}
+	tests := []struct {
+		name string
+		args args
+		want map[uint64][]uint64
+	}{
+		{
+			name: "cartesian join",
+			args: args{
+				seriesTagIndex1: seriesTagIndex1,
+				seriesTagIndex2: seriesTagIndex2,
+			},
+			want: map[uint64][]uint64{
+				0: {1, 2, 3, 4, 9, 10, 11, 12},
+				1: {5, 6, 7, 8, 9, 10, 11, 12},
+				2: {5, 6, 7, 8, 13, 14, 15, 16},
+				3: {1, 2, 3, 4, 13, 14, 15, 16},
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := cartesianJoin(tt.args.seriesTagIndex1, tt.args.seriesTagIndex2); !allValueDeepEqual(got, tt.want) {
+				t.Errorf("cartesianJoin() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_onJoin(t *testing.T) {
+	type args struct {
+		reHashTagIndex1 map[uint64][][]uint64
+		reHashTagIndex2 map[uint64][][]uint64
+		joinType        JoinType
+	}
+	tests := []struct {
+		name string
+		args args
+		want map[uint64][]uint64
+	}{
+		{
+			name: "left join",
+			args: args{
+				reHashTagIndex1: reHashTagIndex1,
+				reHashTagIndex2: reHashTagIndex2,
+				joinType:        Left,
+			},
+			want: map[uint64][]uint64{
+				1: {1, 2, 9, 10},
+				2: {3, 4, 9, 10},
+				3: {1, 2, 11, 12},
+				4: {3, 4, 11, 12},
+				5: {5, 6},
+				6: {7, 8},
+			},
+		},
+		{
+			name: "right join",
+			args: args{
+				reHashTagIndex1: reHashTagIndex2,
+				reHashTagIndex2: reHashTagIndex1,
+				joinType:        Right,
+			},
+			want: map[uint64][]uint64{
+				1: {1, 2, 9, 10},
+				2: {3, 4, 9, 10},
+				3: {1, 2, 11, 12},
+				4: {3, 4, 11, 12},
+				5: {13, 14},
+				6: {15, 16},
+			},
+		},
+
+		{
+			name: "inner join",
+			args: args{
+				reHashTagIndex1: reHashTagIndex1,
+				reHashTagIndex2: reHashTagIndex2,
+				joinType:        Inner,
+			},
+			want: map[uint64][]uint64{
+				1: {1, 2, 9, 10},
+				2: {3, 4, 9, 10},
+				3: {1, 2, 11, 12},
+				4: {3, 4, 11, 12},
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := onJoin(tt.args.reHashTagIndex1, tt.args.reHashTagIndex2, tt.args.joinType); !allValueDeepEqual(flatten(got), tt.want) {
+				t.Errorf("onJoin() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+// allValueDeepEqual 判断 map 的 value 是否相同，不考虑 key
+func allValueDeepEqual(got, want map[uint64][]uint64) bool {
+	if len(got) != len(want) {
+		return false
+	}
+	for _, v1 := range got {
+		curEqual := false
+		slices.Sort(v1)
+		for _, v2 := range want {
+			slices.Sort(v2)
+			if reflect.DeepEqual(v1, v2) {
+				curEqual = true
+				break
+			}
+		}
+		if !curEqual {
+			return false
+		}
+	}
+	return true
+}
+
+// allValueDeepEqualOmitOrder 判断两个字符串切片是否相等，不考虑顺序
+func allValueDeepEqualOmitOrder(got, want []string) bool {
+	if len(got) != len(want) {
+		return false
+	}
+	slices.Sort(got)
+	slices.Sort(want)
+	for i := range got {
+		if got[i] != want[i] {
+			return false
+		}
+	}
+	return true
+}
+
+func Test_removeVal(t *testing.T) {
+	type args struct {
+		promql string
+	}
+	tests := []struct {
+		name string
+		args args
+		want string
+	}{
+		// TODO: Add test cases.
+		{
+			name: "removeVal1",
+			args: args{
+				promql: "mem{test1=\"$test1\",test2=\"$test2\",test3=\"$test3\"} > $val",
+			},
+			want: "mem{} > $val",
+		},
+		{
+			name: "removeVal2",
+			args: args{
+				promql: "mem{test1=\"test1\",test2=\"$test2\",test3=\"$test3\"} > $val",
+			},
+			want: "mem{test1=\"test1\"} > $val",
+		},
+		{
+			name: "removeVal3",
+			args: args{
+				promql: "mem{test1=\"$test1\",test2=\"test2\",test3=\"$test3\"} > $val",
+			},
+			want: "mem{test2=\"test2\"} > $val",
+		},
+		{
+			name: "removeVal4",
+			args: args{
+				promql: "mem{test1=\"$test1\",test2=\"$test2\",test3=\"test3\"} > $val",
+			},
+			want: "mem{test3=\"test3\"} > $val",
+		},
+		{
+			name: "removeVal5",
+			args: args{
+				promql: "mem{test1=\"$test1\",test2=\"test2\",test3=\"test3\"} > $val",
+			},
+			want: "mem{test2=\"test2\",test3=\"test3\"} > $val",
+		},
+		{
+			name: "removeVal6",
+			args: args{
+				promql: "mem{test1=\"test1\",test2=\"$test2\",test3=\"test3\"} > $val",
+			},
+			want: "mem{test1=\"test1\",test3=\"test3\"} > $val",
+		},
+		{
+			name: "removeVal7",
+			args: args{
+				promql: "mem{test1=\"test1\",test2=\"test2\",test3='$test3'} > $val",
+			},
+			want: "mem{test1=\"test1\",test2=\"test2\"} > $val",
+		},
+		{
+			name: "removeVal8",
+			args: args{
+				promql: "mem{test1=\"test1\",test2=\"test2\",test3=\"test3\"} > $val",
+			},
+			want: "mem{test1=\"test1\",test2=\"test2\",test3=\"test3\"} > $val",
+		},
+		{
+			name: "removeVal9",
+			args: args{
+				promql: "mem{test1=\"$test1\",test2=\"test2\"} > $val1 and mem{test3=\"test3\",test4=\"test4\"} > $val2",
+			},
+			want: "mem{test2=\"test2\"} > $val1 and mem{test3=\"test3\",test4=\"test4\"} > $val2",
+		},
+		{
+			name: "removeVal10",
+			args: args{
+				promql: "mem{test1=\"test1\",test2='$test2'} > $val1 and mem{test3=\"test3\",test4=\"test4\"} > $val2",
+			},
+			want: "mem{test1=\"test1\"} > $val1 and mem{test3=\"test3\",test4=\"test4\"} > $val2",
+		},
+		{
+			name: "removeVal11",
+			args: args{
+				promql: "mem{test1='test1',test2=\"test2\"} > $val1 and mem{test3=\"$test3\",test4=\"test4\"} > $val2",
+			},
+			want: "mem{test1='test1',test2=\"test2\"} > $val1 and mem{test4=\"test4\"} > $val2",
+		},
+		{
+			name: "removeVal12",
+			args: args{
+				promql: "mem{test1=\"test1\",test2=\"test2\"} > $val1 and mem{test3=\"test3\",test4=\"$test4\"} > $val2",
+			},
+			want: "mem{test1=\"test1\",test2=\"test2\"} > $val1 and mem{test3=\"test3\"} > $val2",
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := removeVal(tt.args.promql); got != tt.want {
+				t.Errorf("removeVal() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func TestExtractVarMapping(t *testing.T) {
+	tests := []struct {
+		name   string
+		promql string
+		want   map[string]string
+	}{
+		{
+			name:   "单个花括号单个变量",
+			promql: `mem_used_percent{host="$my_host"} > $val`,
+			want:   map[string]string{"my_host": "host"},
+		},
+		{
+			name:   "单个花括号多个变量",
+			promql: `mem_used_percent{host="$my_host",region="$region",env="prod"} > $val`,
+			want:   map[string]string{"my_host": "host", "region": "region"},
+		},
+		{
+			name:   "多个花括号多个变量",
+			promql: `sum(rate(mem_used_percent{host="$my_host"})) by (instance) + avg(node_load1{region="$region"}) > $val`,
+			want:   map[string]string{"my_host": "host", "region": "region"},
+		},
+		{
+			name:   "相同变量出现多次",
+			promql: `sum(rate(mem_used_percent{host="$my_host"})) + avg(node_load1{host="$my_host"}) > $val`,
+			want:   map[string]string{"my_host": "host"},
+		},
+		{
+			name:   "没有变量",
+			promql: `mem_used_percent{host="localhost",region="cn"} > 80`,
+			want:   map[string]string{},
+		},
+		{
+			name:   "没有花括号",
+			promql: `80 > $val`,
+			want:   map[string]string{},
+		},
+		{
+			name:   "格式不规范的标签",
+			promql: `mem_used_percent{host=$my_host,region = $region} > $val`,
+			want:   map[string]string{"my_host": "host", "region": "region"},
+		},
+		{
+			name:   "空花括号",
+			promql: `mem_used_percent{} > $val`,
+			want:   map[string]string{},
+		},
+		{
+			name:   "不完整的花括号",
+			promql: `mem_used_percent{host="$my_host"`,
+			want:   map[string]string{},
+		},
+		{
+			name:   "复杂表达式",
+			promql: `sum(rate(http_requests_total{handler="$handler",code="$code"}[5m])) by (handler) / sum(rate(http_requests_total{handler="$handler"}[5m])) by (handler) * 100 > $threshold`,
+			want:   map[string]string{"handler": "handler", "code": "code"},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := ExtractVarMapping(tt.promql)
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("ExtractVarMapping() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

alert/mute/mute.go

@@ -114,7 +114,7 @@ func BgNotMatchMuteStrategy(rule *models.AlertRule, event *models.AlertCurEvent,
 	target, exists := targetCache.Get(ident)
 	// 对于包含ident的告警事件，check一下ident所属bg和rule所属bg是否相同
 	// 如果告警规则选择了只在本BG生效，那其他BG的机器就不能因此规则产生告警
-	if exists && target.GroupId != rule.GroupId {
+	if exists && !target.MatchGroupId(rule.GroupId) {
 		logger.Debugf("[%s] mute: rule_eval:%d cluster:%s", "BgNotMatchMuteStrategy", rule.Id, event.Cluster)
 		return true
 	}

alert/process/process.go

@@ -2,6 +2,7 @@ package process
 
 import (
 	"bytes"
+	"encoding/json"
 	"fmt"
 	"html/template"
 	"sort"
@@ -21,6 +22,7 @@ import (
 	"github.com/ccfos/nightingale/v6/pushgw/writer"
 
 	"github.com/prometheus/prometheus/prompb"
+	"github.com/robfig/cron/v3"
 	"github.com/toolkits/pkg/logger"
 	"github.com/toolkits/pkg/str"
 )
@@ -53,10 +55,11 @@ type Processor struct {
 	datasourceId int64
 	EngineName   string
 
-	rule     *models.AlertRule
-	fires    *AlertCurEventMap
-	pendings *AlertCurEventMap
-	inhibit  bool
+	rule                 *models.AlertRule
+	fires                *AlertCurEventMap
+	pendings             *AlertCurEventMap
+	pendingsUseByRecover *AlertCurEventMap
+	inhibit              bool
 
 	tagsMap    map[string]string
 	tagsArr    []string
@@ -77,6 +80,9 @@ type Processor struct {
 	HandleFireEventHook    HandleEventFunc
 	HandleRecoverEventHook HandleEventFunc
 	EventMuteHook          EventMuteHookFunc
+
+	ScheduleEntry    cron.Entry
+	PromEvalInterval int
 }
 
 func (p *Processor) Key() string {
@@ -88,9 +94,9 @@ func (p *Processor) DatasourceId() int64 {
 }
 
 func (p *Processor) Hash() string {
-	return str.MD5(fmt.Sprintf("%d_%d_%s_%d",
+	return str.MD5(fmt.Sprintf("%d_%s_%s_%d",
 		p.rule.Id,
-		p.rule.PromEvalInterval,
+		p.rule.CronPattern,
 		p.rule.RuleConfig,
 		p.datasourceId,
 	))
@@ -125,7 +131,7 @@ func NewProcessor(engineName string, rule *models.AlertRule, datasourceId int64,
 	return p
 }
 
-func (p *Processor) Handle(anomalyPoints []common.AnomalyPoint, from string, inhibit bool) {
+func (p *Processor) Handle(anomalyPoints []models.AnomalyPoint, from string, inhibit bool) {
 	// 有可能rule的一些配置已经发生变化，比如告警接收人、callbacks等
 	// 这些信息的修改是不会引起worker restart的，但是确实会影响告警处理逻辑
 	// 所以，这里直接从memsto.AlertRuleCache中获取并覆盖
@@ -133,10 +139,13 @@ func (p *Processor) Handle(anomalyPoints []common.AnomalyPoint, from string, inh
 	cachedRule := p.alertRuleCache.Get(p.rule.Id)
 	if cachedRule == nil {
 		logger.Errorf("rule not found %+v", anomalyPoints)
-		p.Stats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", p.DatasourceId()), "handle_event").Inc()
+		p.Stats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", p.DatasourceId()), "handle_event", p.BusiGroupCache.GetNameByBusiGroupId(p.rule.GroupId), fmt.Sprintf("%v", p.rule.Id)).Inc()
 		return
 	}
 
+	// 在 rule 变化之前取到 ruleHash
+	ruleHash := p.rule.Hash()
+
 	p.rule = cachedRule
 	now := time.Now().Unix()
 	alertingKeys := map[string]struct{}{}
@@ -144,7 +153,7 @@ func (p *Processor) Handle(anomalyPoints []common.AnomalyPoint, from string, inh
 	// 根据 event 的 tag 将 events 分组，处理告警抑制的情况
 	eventsMap := make(map[string][]*models.AlertCurEvent)
 	for _, anomalyPoint := range anomalyPoints {
-		event := p.BuildEvent(anomalyPoint, from, now)
+		event := p.BuildEvent(anomalyPoint, from, now, ruleHash)
 		// 如果 event 被 mute 了,本质也是 fire 的状态,这里无论如何都添加到 alertingKeys 中,防止 fire 的事件自动恢复了
 		hash := event.Hash
 		alertingKeys[hash] = struct{}{}
@@ -169,10 +178,12 @@ func (p *Processor) Handle(anomalyPoints []common.AnomalyPoint, from string, inh
 		p.handleEvent(events)
 	}
 
-	p.HandleRecover(alertingKeys, now, inhibit)
+	if from == "inner" {
+		p.HandleRecover(alertingKeys, now, inhibit)
+	}
 }
 
-func (p *Processor) BuildEvent(anomalyPoint common.AnomalyPoint, from string, now int64) *models.AlertCurEvent {
+func (p *Processor) BuildEvent(anomalyPoint models.AnomalyPoint, from string, now int64, ruleHash string) *models.AlertCurEvent {
 	p.fillTags(anomalyPoint)
 	p.mayHandleIdent()
 	hash := Hash(p.rule.Id, p.datasourceId, anomalyPoint)
@@ -198,18 +209,43 @@ func (p *Processor) BuildEvent(anomalyPoint common.AnomalyPoint, from string, no
 	event.TargetNote = p.targetNote
 	event.TriggerValue = anomalyPoint.ReadableValue()
 	event.TriggerValues = anomalyPoint.Values
+	event.TriggerValuesJson = models.EventTriggerValues{ValuesWithUnit: anomalyPoint.ValuesUnit}
 	event.TagsJSON = p.tagsArr
 	event.Tags = strings.Join(p.tagsArr, ",,")
 	event.IsRecovered = false
 	event.Callbacks = p.rule.Callbacks
 	event.CallbacksJSON = p.rule.CallbacksJSON
 	event.Annotations = p.rule.Annotations
-	event.AnnotationsJSON = make(map[string]string)
 	event.RuleConfig = p.rule.RuleConfig
 	event.RuleConfigJson = p.rule.RuleConfigJson
 	event.Severity = anomalyPoint.Severity
 	event.ExtraConfig = p.rule.ExtraConfigJSON
 	event.PromQl = anomalyPoint.Query
+	event.RecoverConfig = anomalyPoint.RecoverConfig
+	event.RuleHash = ruleHash
+
+	if anomalyPoint.TriggerType == models.TriggerTypeNodata {
+		event.TriggerValue = "nodata"
+		ruleConfig := models.RuleQuery{}
+		json.Unmarshal([]byte(p.rule.RuleConfig), &ruleConfig)
+		ruleConfig.TriggerType = anomalyPoint.TriggerType
+		b, _ := json.Marshal(ruleConfig)
+		event.RuleConfig = string(b)
+	}
+
+	if err := json.Unmarshal([]byte(p.rule.Annotations), &event.AnnotationsJSON); err != nil {
+		event.AnnotationsJSON = make(map[string]string) // 解析失败时使用空 map
+		logger.Warningf("unmarshal annotations json failed: %v, rule: %d", err, p.rule.Id)
+	}
+
+	if p.target != "" {
+		if pt, exist := p.TargetCache.Get(p.target); exist {
+			pt.GroupNames = p.BusiGroupCache.GetNamesByBusiGroupIds(pt.GroupIds)
+			event.Target = pt
+		} else {
+			logger.Infof("Target[ident: %s] doesn't exist in cache.", p.target)
+		}
+	}
 
 	if event.TriggerValues != "" && strings.Count(event.TriggerValues, "$") > 1 {
 		// TriggerValues 有多个变量，将多个变量都放到 TriggerValue 中
@@ -232,17 +268,17 @@ func Relabel(rule *models.AlertRule, event *models.AlertCurEvent) {
 		return
 	}
 
+	if len(rule.EventRelabelConfig) == 0 {
+		return
+	}
+
 	// need to keep the original label
 	event.OriginalTags = event.Tags
 	event.OriginalTagsJSON = make([]string, len(event.TagsJSON))
 
 	labels := make([]prompb.Label, len(event.TagsJSON))
 	for i, tag := range event.TagsJSON {
-		label := strings.Split(tag, "=")
-		if len(label) != 2 {
-			logger.Errorf("event%+v relabel: the label length is not 2:%v", event, label)
-			continue
-		}
+		label := strings.SplitN(tag, "=", 2)
 		event.OriginalTagsJSON[i] = tag
 		labels[i] = prompb.Label{Name: label[0], Value: label[1]}
 	}
@@ -281,7 +317,7 @@ func (p *Processor) HandleRecover(alertingKeys map[string]struct{}, now int64, i
 	}
 
 	hashArr := make([]string, 0, len(alertingKeys))
-	for hash := range p.fires.GetAll() {
+	for hash, _ := range p.fires.GetAll() {
 		if _, has := alertingKeys[hash]; has {
 			continue
 		}
@@ -300,7 +336,7 @@ func (p *Processor) HandleRecoverEvent(hashArr []string, now int64, inhibit bool
 
 	if !inhibit {
 		for _, hash := range hashArr {
-			p.RecoverSingle(hash, now, nil)
+			p.RecoverSingle(false, hash, now, nil)
 		}
 		return
 	}
@@ -328,11 +364,11 @@ func (p *Processor) HandleRecoverEvent(hashArr []string, now int64, inhibit bool
 	}
 
 	for _, event := range eventMap {
-		p.RecoverSingle(event.Hash, now, nil)
+		p.RecoverSingle(false, event.Hash, now, nil)
 	}
 }
 
-func (p *Processor) RecoverSingle(hash string, now int64, value *string, values ...string) {
+func (p *Processor) RecoverSingle(byRecover bool, hash string, now int64, value *string, values ...string) {
 	cachedRule := p.rule
 	if cachedRule == nil {
 		return
@@ -342,11 +378,28 @@ func (p *Processor) RecoverSingle(hash string, now int64, value *string, values
 	if !has {
 		return
 	}
+
 	// 如果配置了留观时长，就不能立马恢复了
-	if cachedRule.RecoverDuration > 0 && now-event.LastEvalTime < cachedRule.RecoverDuration {
+	if cachedRule.RecoverDuration > 0 {
+		lastPendingEvent, has := p.pendingsUseByRecover.Get(hash)
+		if !has {
+			// 说明没有产生过异常点，就不需要恢复了
+			logger.Debugf("rule_eval:%s event:%v do not has pending event, not recover", p.Key(), event)
+			return
+		}
+
+		if now-lastPendingEvent.LastEvalTime < cachedRule.RecoverDuration {
+			logger.Debugf("rule_eval:%s event:%v not recover", p.Key(), event)
+			return
+		}
+	}
+
+	// 如果设置了恢复条件，则不能在此处恢复，必须依靠 recoverPoint 来恢复
+	if event.RecoverConfig.JudgeType != models.Origin && !byRecover {
 		logger.Debugf("rule_eval:%s event:%v not recover", p.Key(), event)
 		return
 	}
+
 	if value != nil {
 		event.TriggerValue = *value
 		if len(values) > 0 {
@@ -358,6 +411,7 @@ func (p *Processor) RecoverSingle(hash string, now int64, value *string, values
 	// 我确实无法分辨，是prom中有值但是未满足阈值所以没返回，还是prom中确实丢了一些点导致没有数据可以返回，尴尬
 	p.fires.Delete(hash)
 	p.pendings.Delete(hash)
+	p.pendingsUseByRecover.Delete(hash)
 
 	// 可能是因为调整了promql才恢复的，所以事件里边要体现最新的promql，否则用户会比较困惑
 	// 当然，其实rule的各个字段都可能发生变化了，都更新一下吧
@@ -377,6 +431,14 @@ func (p *Processor) handleEvent(events []*models.AlertCurEvent) {
 		if event == nil {
 			continue
 		}
+
+		if _, has := p.pendingsUseByRecover.Get(event.Hash); has {
+			p.pendingsUseByRecover.UpdateLastEvalTime(event.Hash, event.LastEvalTime)
+		} else {
+			p.pendingsUseByRecover.Set(event.Hash, event)
+		}
+
+		event.PromEvalInterval = p.PromEvalInterval
 		if p.rule.PromForDuration == 0 {
 			fireEvents = append(fireEvents, event)
 			if severity > event.Severity {
@@ -385,7 +447,7 @@ func (p *Processor) handleEvent(events []*models.AlertCurEvent) {
 			continue
 		}
 
-		var preTriggerTime int64
+		var preTriggerTime int64 // 第一个 pending event 的触发时间
 		preEvent, has := p.pendings.Get(event.Hash)
 		if has {
 			p.pendings.UpdateLastEvalTime(event.Hash, event.LastEvalTime)
@@ -471,22 +533,24 @@ func (p *Processor) pushEventToQueue(e *models.AlertCurEvent) {
 	dispatch.LogEvent(e, "push_queue")
 	if !queue.EventQueue.PushFront(e) {
 		logger.Warningf("event_push_queue: queue is full, event:%+v", e)
-		p.Stats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", p.DatasourceId()), "push_event_queue").Inc()
+		p.Stats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", p.DatasourceId()), "push_event_queue", p.BusiGroupCache.GetNameByBusiGroupId(p.rule.GroupId), fmt.Sprintf("%v", p.rule.Id)).Inc()
 	}
 }
 
 func (p *Processor) RecoverAlertCurEventFromDb() {
 	p.pendings = NewAlertCurEventMap(nil)
+	p.pendingsUseByRecover = NewAlertCurEventMap(nil)
 
 	curEvents, err := models.AlertCurEventGetByRuleIdAndDsId(p.ctx, p.rule.Id, p.datasourceId)
 	if err != nil {
 		logger.Errorf("recover event from db for rule:%s failed, err:%s", p.Key(), err)
-		p.Stats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", p.DatasourceId()), "get_recover_event").Inc()
+		p.Stats.CounterRuleEvalErrorTotal.WithLabelValues(fmt.Sprintf("%v", p.DatasourceId()), "get_recover_event", p.BusiGroupCache.GetNameByBusiGroupId(p.rule.GroupId), fmt.Sprintf("%v", p.rule.Id)).Inc()
 		p.fires = NewAlertCurEventMap(nil)
 		return
 	}
 
 	fireMap := make(map[string]*models.AlertCurEvent)
+	pendingsUseByRecoverMap := make(map[string]*models.AlertCurEvent)
 	for _, event := range curEvents {
 		if event.Cate == models.HOST {
 			target, exists := p.TargetCache.Get(event.TargetIdent)
@@ -497,13 +561,24 @@ func (p *Processor) RecoverAlertCurEventFromDb() {
 		}
 
 		event.DB2Mem()
+		target, exists := p.TargetCache.Get(event.TargetIdent)
+		if exists {
+			target.GroupNames = p.BusiGroupCache.GetNamesByBusiGroupIds(target.GroupIds)
+			event.Target = target
+		}
+
 		fireMap[event.Hash] = event
+		e := *event
+		pendingsUseByRecoverMap[event.Hash] = &e
 	}
 
 	p.fires = NewAlertCurEventMap(fireMap)
+
+	// 修改告警规则，或者进程重启之后，需要重新加载 pendingsUseByRecover
+	p.pendingsUseByRecover = NewAlertCurEventMap(pendingsUseByRecoverMap)
 }
 
-func (p *Processor) fillTags(anomalyPoint common.AnomalyPoint) {
+func (p *Processor) fillTags(anomalyPoint models.AnomalyPoint) {
 	// handle series tags
 	tagsMap := make(map[string]string)
 	for label, value := range anomalyPoint.Labels {
@@ -576,6 +651,7 @@ func (p *Processor) mayHandleGroup() {
 func (p *Processor) DeleteProcessEvent(hash string) {
 	p.fires.Delete(hash)
 	p.pendings.Delete(hash)
+	p.pendingsUseByRecover.Delete(hash)
 }
 
 func labelMapToArr(m map[string]string) []string {
@@ -592,10 +668,10 @@ func labelMapToArr(m map[string]string) []string {
 	return labelStrings
 }
 
-func Hash(ruleId, datasourceId int64, vector common.AnomalyPoint) string {
+func Hash(ruleId, datasourceId int64, vector models.AnomalyPoint) string {
 	return str.MD5(fmt.Sprintf("%d_%s_%d_%d_%s", ruleId, vector.Labels.String(), datasourceId, vector.Severity, vector.Query))
 }
 
-func TagHash(vector common.AnomalyPoint) string {
+func TagHash(vector models.AnomalyPoint) string {
 	return str.MD5(vector.Labels.String())
 }

alert/record/prom_rule.go

@@ -56,11 +56,12 @@ func (rrc *RecordRuleContext) Key() string {
 }
 
 func (rrc *RecordRuleContext) Hash() string {
-	return str.MD5(fmt.Sprintf("%d_%s_%s_%d",
+	return str.MD5(fmt.Sprintf("%d_%s_%s_%d_%s",
 		rrc.rule.Id,
 		rrc.rule.CronPattern,
 		rrc.rule.PromQl,
 		rrc.datasourceId,
+		rrc.rule.AppendTags,
 	))
 }
 

alert/record/scheduler.go

@@ -26,9 +26,11 @@ type Scheduler struct {
 	writers     *writer.WritersType
 
 	stats *astats.Stats
+
+	datasourceCache *memsto.DatasourceCacheType
 }
 
-func NewScheduler(aconf aconf.Alert, rrc *memsto.RecordingRuleCacheType, promClients *prom.PromClientMap, writers *writer.WritersType, stats *astats.Stats) *Scheduler {
+func NewScheduler(aconf aconf.Alert, rrc *memsto.RecordingRuleCacheType, promClients *prom.PromClientMap, writers *writer.WritersType, stats *astats.Stats, datasourceCache *memsto.DatasourceCacheType) *Scheduler {
 	scheduler := &Scheduler{
 		aconf:       aconf,
 		recordRules: make(map[string]*RecordRuleContext),
@@ -39,6 +41,8 @@ func NewScheduler(aconf aconf.Alert, rrc *memsto.RecordingRuleCacheType, promCli
 		writers:     writers,
 
 		stats: stats,
+
+		datasourceCache: datasourceCache,
 	}
 
 	go scheduler.LoopSyncRules(context.Background())
@@ -67,7 +71,7 @@ func (s *Scheduler) syncRecordRules() {
 			continue
 		}
 
-		datasourceIds := s.promClients.Hit(rule.DatasourceIdsJson)
+		datasourceIds := s.datasourceCache.GetIDsByDsCateAndQueries("prometheus", rule.DatasourceQueries)
 		for _, dsId := range datasourceIds {
 			if !naming.DatasourceHashRing.IsHit(strconv.FormatInt(dsId, 10), fmt.Sprintf("%d", rule.Id), s.aconf.Heartbeat.Endpoint) {
 				continue

alert/router/router_event.go

@@ -6,7 +6,6 @@ import (
 	"strings"
 	"time"
 
-	"github.com/ccfos/nightingale/v6/alert/common"
 	"github.com/ccfos/nightingale/v6/alert/dispatch"
 	"github.com/ccfos/nightingale/v6/alert/mute"
 	"github.com/ccfos/nightingale/v6/alert/naming"
@@ -34,7 +33,7 @@ func (rt *Router) pushEventToQueue(c *gin.Context) {
 			continue
 		}
 
-		arr := strings.Split(pair, "=")
+		arr := strings.SplitN(pair, "=", 2)
 		if len(arr) != 2 {
 			continue
 		}
@@ -92,7 +91,7 @@ func (rt *Router) eventPersist(c *gin.Context) {
 
 type eventForm struct {
 	Alert         bool                  `json:"alert"`
-	AnomalyPoints []common.AnomalyPoint `json:"vectors"`
+	AnomalyPoints []models.AnomalyPoint `json:"vectors"`
 	RuleId        int64                 `json:"rule_id"`
 	DatasourceId  int64                 `json:"datasource_id"`
 	Inhibit       bool                  `json:"inhibit"`
@@ -129,7 +128,7 @@ func (rt *Router) makeEvent(c *gin.Context) {
 		} else {
 			for _, vector := range events[i].AnomalyPoints {
 				readableString := vector.ReadableValue()
-				go ruleWorker.RecoverSingle(process.Hash(events[i].RuleId, events[i].DatasourceId, vector), vector.Timestamp, &readableString)
+				go ruleWorker.RecoverSingle(false, process.Hash(events[i].RuleId, events[i].DatasourceId, vector), vector.Timestamp, &readableString)
 			}
 		}
 	}

alert/sender/callback.go

@@ -29,13 +29,14 @@ type (
 		Rule        *models.AlertRule
 		Events      []*models.AlertCurEvent
 		Stats       *astats.Stats
+		BatchSend   bool
 	}
 
 	DefaultCallBacker struct{}
 )
 
 func BuildCallBackContext(ctx *ctx.Context, callBackURL string, rule *models.AlertRule, events []*models.AlertCurEvent,
-	uids []int64, userCache *memsto.UserCacheType, stats *astats.Stats) CallBackContext {
+	uids []int64, userCache *memsto.UserCacheType, batchSend bool, stats *astats.Stats) CallBackContext {
 	users := userCache.GetByUserIds(uids)
 
 	newCallBackUrl, _ := events[0].ParseURL(callBackURL)
@@ -45,6 +46,7 @@ func BuildCallBackContext(ctx *ctx.Context, callBackURL string, rule *models.Ale
 		Rule:        rule,
 		Events:      events,
 		Users:       users,
+		BatchSend:   batchSend,
 		Stats:       stats,
 	}
 }
@@ -112,31 +114,96 @@ func (c *DefaultCallBacker) CallBack(ctx CallBackContext) {
 
 	event := ctx.Events[0]
 
-	ctx.Stats.AlertNotifyTotal.WithLabelValues("rule_callback").Inc()
-	resp, code, err := poster.PostJSON(ctx.CallBackURL, 5*time.Second, event, 3)
-	if err != nil {
-		logger.Errorf("event_callback_fail(rule_id=%d url=%s), event:%+v, resp: %s, err: %v, code: %d",
-			event.RuleId, ctx.CallBackURL, event, string(resp), err, code)
-		ctx.Stats.AlertNotifyErrorTotal.WithLabelValues("rule_callback").Inc()
-	} else {
-		logger.Infof("event_callback_succ(rule_id=%d url=%s), event:%+v, resp: %s, code: %d",
-			event.RuleId, ctx.CallBackURL, event, string(resp), code)
+	if ctx.BatchSend {
+		webhookConf := &models.Webhook{
+			Type:          models.RuleCallback,
+			Enable:        true,
+			Url:           ctx.CallBackURL,
+			Timeout:       5,
+			RetryCount:    3,
+			RetryInterval: 10,
+			Batch:         1000,
+		}
+
+		PushCallbackEvent(ctx.Ctx, webhookConf, event, ctx.Stats)
+		return
+	}
+
+	doSendAndRecord(ctx.Ctx, ctx.CallBackURL, ctx.CallBackURL, event, "callback", ctx.Stats, ctx.Events)
+}
+
+func doSendAndRecord(ctx *ctx.Context, url, token string, body interface{}, channel string,
+	stats *astats.Stats, events []*models.AlertCurEvent) {
+	res, err := doSend(url, body, channel, stats)
+	NotifyRecord(ctx, events, channel, token, res, err)
+}
+
+func NotifyRecord(ctx *ctx.Context, evts []*models.AlertCurEvent, channel, target, res string, err error) {
+	// 一个通知可能对应多个 event，都需要记录
+	notis := make([]*models.NotificaitonRecord, 0, len(evts))
+	for _, evt := range evts {
+		noti := models.NewNotificationRecord(evt, channel, target)
+		if err != nil {
+			noti.SetStatus(models.NotiStatusFailure)
+			noti.SetDetails(err.Error())
+		} else if res != "" {
+			noti.SetDetails(string(res))
+		}
+		notis = append(notis, noti)
+	}
+
+	if !ctx.IsCenter {
+		_, err := poster.PostByUrlsWithResp[[]int64](ctx, "/v1/n9e/notify-record", notis)
+		if err != nil {
+			logger.Errorf("add notis:%v failed, err: %v", notis, err)
+		}
+		return
+	}
+
+	if err := models.DB(ctx).CreateInBatches(notis, 100).Error; err != nil {
+		logger.Errorf("add notis:%v failed, err: %v", notis, err)
 	}
 }
 
-func doSend(url string, body interface{}, channel string, stats *astats.Stats) {
+func doSend(url string, body interface{}, channel string, stats *astats.Stats) (string, error) {
 	stats.AlertNotifyTotal.WithLabelValues(channel).Inc()
 
 	res, code, err := poster.PostJSON(url, time.Second*5, body, 3)
 	if err != nil {
 		logger.Errorf("%s_sender: result=fail url=%s code=%d error=%v req:%v response=%s", channel, url, code, err, body, string(res))
 		stats.AlertNotifyErrorTotal.WithLabelValues(channel).Inc()
-	} else {
-		logger.Infof("%s_sender: result=succ url=%s code=%d req:%v response=%s", channel, url, code, body, string(res))
+		return "", err
 	}
+
+	logger.Infof("%s_sender: result=succ url=%s code=%d req:%v response=%s", channel, url, code, body, string(res))
+	return string(res), nil
 }
 
 type TaskCreateReply struct {
 	Err string `json:"err"`
 	Dat int64  `json:"dat"` // task.id
 }
+
+func PushCallbackEvent(ctx *ctx.Context, webhook *models.Webhook, event *models.AlertCurEvent, stats *astats.Stats) {
+	CallbackEventQueueLock.RLock()
+	queue := CallbackEventQueue[webhook.Url]
+	CallbackEventQueueLock.RUnlock()
+
+	if queue == nil {
+		queue = &WebhookQueue{
+			eventQueue: NewSafeEventQueue(QueueMaxSize),
+			closeCh:    make(chan struct{}),
+		}
+
+		CallbackEventQueueLock.Lock()
+		CallbackEventQueue[webhook.Url] = queue
+		CallbackEventQueueLock.Unlock()
+
+		StartConsumer(ctx, queue, webhook.Batch, webhook, stats)
+	}
+
+	succ := queue.eventQueue.Push(event)
+	if !succ {
+		logger.Warningf("Write channel(%s) full, current channel size: %d event:%v", webhook.Url, queue.eventQueue.Len(), event)
+	}
+}

alert/sender/dingtalk.go

@@ -1,9 +1,10 @@
 package sender
 
 import (
-	"github.com/ccfos/nightingale/v6/models"
 	"html/template"
 	"strings"
+
+	"github.com/ccfos/nightingale/v6/models"
 )
 
 type dingtalkMarkdown struct {
@@ -35,13 +36,13 @@ func (ds *DingtalkSender) Send(ctx MessageContext) {
 		return
 	}
 
-	urls, ats := ds.extract(ctx.Users)
+	urls, ats, tokens := ds.extract(ctx.Users)
 	if len(urls) == 0 {
 		return
 	}
 	message := BuildTplMessage(models.Dingtalk, ds.tpl, ctx.Events)
 
-	for _, url := range urls {
+	for i, url := range urls {
 		var body dingtalk
 		// NoAt in url
 		if strings.Contains(url, "noat=1") {
@@ -66,7 +67,7 @@ func (ds *DingtalkSender) Send(ctx MessageContext) {
 			}
 		}
 
-		doSend(url, body, models.Dingtalk, ctx.Stats)
+		doSendAndRecord(ctx.Ctx, url, tokens[i], body, models.Dingtalk, ctx.Stats, ctx.Events)
 	}
 }
 
@@ -96,15 +97,14 @@ func (ds *DingtalkSender) CallBack(ctx CallBackContext) {
 		body.Markdown.Text = message
 	}
 
-	doSend(ctx.CallBackURL, body, models.Dingtalk, ctx.Stats)
-
-	ctx.Stats.AlertNotifyTotal.WithLabelValues("rule_callback").Inc()
+	doSendAndRecord(ctx.Ctx, ctx.CallBackURL, ctx.CallBackURL, body, "callback", ctx.Stats, ctx.Events)
 }
 
 // extract urls and ats from Users
-func (ds *DingtalkSender) extract(users []*models.User) ([]string, []string) {
+func (ds *DingtalkSender) extract(users []*models.User) ([]string, []string, []string) {
 	urls := make([]string, 0, len(users))
 	ats := make([]string, 0, len(users))
+	tokens := make([]string, 0, len(users))
 
 	for _, user := range users {
 		if user.Phone != "" {
@@ -116,7 +116,8 @@ func (ds *DingtalkSender) extract(users []*models.User) ([]string, []string) {
 				url = "https://oapi.dingtalk.com/robot/send?access_token=" + token
 			}
 			urls = append(urls, url)
+			tokens = append(tokens, token)
 		}
 	}
-	return urls, ats
+	return urls, ats, tokens
 }

alert/sender/email.go

@@ -9,13 +9,14 @@ import (
 	"github.com/ccfos/nightingale/v6/alert/aconf"
 	"github.com/ccfos/nightingale/v6/memsto"
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
 
 	"github.com/toolkits/pkg/logger"
 
 	"gopkg.in/gomail.v2"
 )
 
-var mailch chan *gomail.Message
+var mailch chan *EmailContext
 
 type EmailSender struct {
 	subjectTpl *template.Template
@@ -23,6 +24,11 @@ type EmailSender struct {
 	smtp       aconf.SMTPConfig
 }
 
+type EmailContext struct {
+	events []*models.AlertCurEvent
+	mail   *gomail.Message
+}
+
 func (es *EmailSender) Send(ctx MessageContext) {
 	if len(ctx.Users) == 0 || len(ctx.Events) == 0 {
 		return
@@ -36,7 +42,7 @@ func (es *EmailSender) Send(ctx MessageContext) {
 		subject = ctx.Events[0].RuleName
 	}
 	content := BuildTplMessage(models.Email, es.contentTpl, ctx.Events)
-	es.WriteEmail(subject, content, tos)
+	es.WriteEmail(subject, content, tos, ctx.Events)
 
 	ctx.Stats.AlertNotifyTotal.WithLabelValues(models.Email).Add(float64(len(tos)))
 }
@@ -73,7 +79,7 @@ func SendEmail(subject, content string, tos []string, stmp aconf.SMTPConfig) err
 	return nil
 }
 
-func (es *EmailSender) WriteEmail(subject, content string, tos []string) {
+func (es *EmailSender) WriteEmail(subject, content string, tos []string, events []*models.AlertCurEvent) {
 	m := gomail.NewMessage()
 
 	m.SetHeader("From", es.smtp.From)
@@ -81,7 +87,7 @@ func (es *EmailSender) WriteEmail(subject, content string, tos []string) {
 	m.SetHeader("Subject", subject)
 	m.SetBody("text/html", content)
 
-	mailch <- m
+	mailch <- &EmailContext{events, m}
 }
 
 func dialSmtp(d *gomail.Dialer) gomail.SendCloser {
@@ -104,22 +110,22 @@ func dialSmtp(d *gomail.Dialer) gomail.SendCloser {
 
 var mailQuit = make(chan struct{})
 
-func RestartEmailSender(smtp aconf.SMTPConfig) {
+func RestartEmailSender(ctx *ctx.Context, smtp aconf.SMTPConfig) {
 	// Notify internal start exit
 	mailQuit <- struct{}{}
-	startEmailSender(smtp)
+	startEmailSender(ctx, smtp)
 }
 
 var smtpConfig aconf.SMTPConfig
 
-func InitEmailSender(ncc *memsto.NotifyConfigCacheType) {
-	mailch = make(chan *gomail.Message, 100000)
-	go updateSmtp(ncc)
+func InitEmailSender(ctx *ctx.Context, ncc *memsto.NotifyConfigCacheType) {
+	mailch = make(chan *EmailContext, 100000)
+	go updateSmtp(ctx, ncc)
 	smtpConfig = ncc.GetSMTP()
-	startEmailSender(smtpConfig)
+	go startEmailSender(ctx, smtpConfig)
 }
 
-func updateSmtp(ncc *memsto.NotifyConfigCacheType) {
+func updateSmtp(ctx *ctx.Context, ncc *memsto.NotifyConfigCacheType) {
 	for {
 		time.Sleep(1 * time.Minute)
 		smtp := ncc.GetSMTP()
@@ -127,15 +133,16 @@ func updateSmtp(ncc *memsto.NotifyConfigCacheType) {
 			smtpConfig.Pass != smtp.Pass || smtpConfig.User != smtp.User || smtpConfig.Port != smtp.Port ||
 			smtpConfig.InsecureSkipVerify != smtp.InsecureSkipVerify { //diff
 			smtpConfig = smtp
-			RestartEmailSender(smtp)
+			RestartEmailSender(ctx, smtp)
 		}
 	}
 }
 
-func startEmailSender(smtp aconf.SMTPConfig) {
+func startEmailSender(ctx *ctx.Context, smtp aconf.SMTPConfig) {
 	conf := smtp
 	if conf.Host == "" || conf.Port == 0 {
 		logger.Warning("SMTP configurations invalid")
+		<-mailQuit
 		return
 	}
 	logger.Infof("start email sender... conf.Host:%+v,conf.Port:%+v", conf.Host, conf.Port)
@@ -167,7 +174,8 @@ func startEmailSender(smtp aconf.SMTPConfig) {
 				}
 				open = true
 			}
-			if err := gomail.Send(s, m); err != nil {
+			var err error
+			if err = gomail.Send(s, m.mail); err != nil {
 				logger.Errorf("email_sender: failed to send: %s", err)
 
 				// close and retry
@@ -184,11 +192,20 @@ func startEmailSender(smtp aconf.SMTPConfig) {
 				}
 				open = true
 
-				if err := gomail.Send(s, m); err != nil {
+				if err = gomail.Send(s, m.mail); err != nil {
 					logger.Errorf("email_sender: failed to retry send: %s", err)
 				}
 			} else {
-				logger.Infof("email_sender: result=succ subject=%v to=%v", m.GetHeader("Subject"), m.GetHeader("To"))
+				logger.Infof("email_sender: result=succ subject=%v to=%v",
+					m.mail.GetHeader("Subject"), m.mail.GetHeader("To"))
+			}
+
+			for _, to := range m.mail.GetHeader("To") {
+				msg := ""
+				if err == nil {
+					msg = "ok"
+				}
+				NotifyRecord(ctx, m.events, models.Email, to, msg, err)
 			}
 
 			size++

alert/sender/feishu.go

@@ -54,17 +54,16 @@ func (fs *FeishuSender) CallBack(ctx CallBackContext) {
 		},
 	}
 
-	doSend(ctx.CallBackURL, body, models.Feishu, ctx.Stats)
-	ctx.Stats.AlertNotifyTotal.WithLabelValues("rule_callback").Inc()
+	doSendAndRecord(ctx.Ctx, ctx.CallBackURL, ctx.CallBackURL, body, "callback", ctx.Stats, ctx.Events)
 }
 
 func (fs *FeishuSender) Send(ctx MessageContext) {
 	if len(ctx.Users) == 0 || len(ctx.Events) == 0 {
 		return
 	}
-	urls, ats := fs.extract(ctx.Users)
+	urls, ats, tokens := fs.extract(ctx.Users)
 	message := BuildTplMessage(models.Feishu, fs.tpl, ctx.Events)
-	for _, url := range urls {
+	for i, url := range urls {
 		body := feishu{
 			Msgtype: "text",
 			Content: feishuContent{
@@ -77,13 +76,14 @@ func (fs *FeishuSender) Send(ctx MessageContext) {
 				IsAtAll:   false,
 			}
 		}
-		doSend(url, body, models.Feishu, ctx.Stats)
+		doSendAndRecord(ctx.Ctx, url, tokens[i], body, models.Feishu, ctx.Stats, ctx.Events)
 	}
 }
 
-func (fs *FeishuSender) extract(users []*models.User) ([]string, []string) {
+func (fs *FeishuSender) extract(users []*models.User) ([]string, []string, []string) {
 	urls := make([]string, 0, len(users))
 	ats := make([]string, 0, len(users))
+	tokens := make([]string, 0, len(users))
 
 	for _, user := range users {
 		if user.Phone != "" {
@@ -95,7 +95,8 @@ func (fs *FeishuSender) extract(users []*models.User) ([]string, []string) {
 				url = "https://open.feishu.cn/open-apis/bot/v2/hook/" + token
 			}
 			urls = append(urls, url)
+			tokens = append(tokens, token)
 		}
 	}
-	return urls, ats
+	return urls, ats, tokens
 }

alert/sender/feishucard.go

@@ -56,8 +56,8 @@ const (
 	Triggered = "triggered"
 )
 
-var (
-	body = feishuCard{
+func createFeishuCardBody() feishuCard {
+	return feishuCard{
 		feishu: feishu{Msgtype: "interactive"},
 		Card: Cards{
 			Config: Conf{
@@ -90,7 +90,7 @@ var (
 			},
 		},
 	}
-)
+}
 
 func (fs *FeishuCardSender) CallBack(ctx CallBackContext) {
 	if len(ctx.Events) == 0 || len(ctx.CallBackURL) == 0 {
@@ -121,6 +121,7 @@ func (fs *FeishuCardSender) CallBack(ctx CallBackContext) {
 	}
 
 	SendTitle := fmt.Sprintf("🔔 %s", ctx.Events[0].RuleName)
+	body := createFeishuCardBody()
 	body.Card.Header.Title.Content = SendTitle
 	body.Card.Header.Template = color
 	body.Card.Elements[0].Text.Content = message
@@ -134,14 +135,14 @@ func (fs *FeishuCardSender) CallBack(ctx CallBackContext) {
 	}
 	parsedURL.RawQuery = ""
 
-	doSend(parsedURL.String(), body, models.FeishuCard, ctx.Stats)
+	doSendAndRecord(ctx.Ctx, parsedURL.String(), parsedURL.String(), body, "callback", ctx.Stats, ctx.Events)
 }
 
 func (fs *FeishuCardSender) Send(ctx MessageContext) {
 	if len(ctx.Users) == 0 || len(ctx.Events) == 0 {
 		return
 	}
-	urls, _ := fs.extract(ctx.Users)
+	urls, tokens := fs.extract(ctx.Users)
 	message := BuildTplMessage(models.FeishuCard, fs.tpl, ctx.Events)
 	color := "red"
 	lowerUnicode := strings.ToLower(message)
@@ -152,18 +153,19 @@ func (fs *FeishuCardSender) Send(ctx MessageContext) {
 	}
 
 	SendTitle := fmt.Sprintf("🔔 %s", ctx.Events[0].RuleName)
+	body := createFeishuCardBody()
 	body.Card.Header.Title.Content = SendTitle
 	body.Card.Header.Template = color
 	body.Card.Elements[0].Text.Content = message
 	body.Card.Elements[2].Elements[0].Content = SendTitle
-	for _, url := range urls {
-		doSend(url, body, models.FeishuCard, ctx.Stats)
+	for i, url := range urls {
+		doSendAndRecord(ctx.Ctx, url, tokens[i], body, models.FeishuCard, ctx.Stats, ctx.Events)
 	}
 }
 
 func (fs *FeishuCardSender) extract(users []*models.User) ([]string, []string) {
 	urls := make([]string, 0, len(users))
-	ats := make([]string, 0)
+	tokens := make([]string, 0, len(users))
 	for i := range users {
 		if token, has := users[i].ExtractToken(models.FeishuCard); has {
 			url := token
@@ -171,7 +173,8 @@ func (fs *FeishuCardSender) extract(users []*models.User) ([]string, []string) {
 				url = "https://open.feishu.cn/open-apis/bot/v2/hook/" + strings.TrimSpace(token)
 			}
 			urls = append(urls, url)
+			tokens = append(tokens, token)
 		}
 	}
-	return urls, ats
+	return urls, tokens
 }

alert/sender/ibex.go

@@ -77,15 +77,20 @@ func (c *IbexCallBacker) handleIbex(ctx *ctx.Context, url string, event *models.
 		return
 	}
 
-	tpl := c.taskTplCache.Get(id)
+	CallIbex(ctx, id, host, c.taskTplCache, c.targetCache, c.userCache, event)
+}
+
+func CallIbex(ctx *ctx.Context, id int64, host string,
+	taskTplCache *memsto.TaskTplCache, targetCache *memsto.TargetCacheType,
+	userCache *memsto.UserCacheType, event *models.AlertCurEvent) {
+	tpl := taskTplCache.Get(id)
 	if tpl == nil {
 		logger.Errorf("event_callback_ibex: no such tpl(%d)", id)
 		return
 	}
-
 	// check perm
 	// tpl.GroupId - host - account 三元组校验权限
-	can, err := canDoIbex(tpl.UpdateBy, tpl, host, c.targetCache, c.userCache)
+	can, err := canDoIbex(tpl.UpdateBy, tpl, host, targetCache, userCache)
 	if err != nil {
 		logger.Errorf("event_callback_ibex: check perm fail: %v", err)
 		return
@@ -103,7 +108,7 @@ func (c *IbexCallBacker) handleIbex(ctx *ctx.Context, url string, event *models.
 			continue
 		}
 
-		arr := strings.Split(pair, "=")
+		arr := strings.SplitN(pair, "=", 2)
 		if len(arr) != 2 {
 			continue
 		}
@@ -113,6 +118,7 @@ func (c *IbexCallBacker) handleIbex(ctx *ctx.Context, url string, event *models.
 	// 附加告警级别  告警触发值标签
 	tagsMap["alert_severity"] = strconv.Itoa(event.Severity)
 	tagsMap["alert_trigger_value"] = event.TriggerValue
+	tagsMap["is_recovered"] = strconv.FormatBool(event.IsRecovered)
 
 	tags, err := json.Marshal(tagsMap)
 	if err != nil {
@@ -176,10 +182,15 @@ func canDoIbex(username string, tpl *models.TaskTpl, host string, targetCache *m
 		return false, nil
 	}
 
-	return target.GroupId == tpl.GroupId, nil
+	return target.MatchGroupId(tpl.GroupId), nil
 }
 
 func TaskAdd(f models.TaskForm, authUser string, isCenter bool) (int64, error) {
+	if storage.Cache == nil {
+		logger.Warning("event_callback_ibex: redis cache is nil")
+		return 0, fmt.Errorf("redis cache is nil")
+	}
+
 	hosts := cleanHosts(f.Hosts)
 	if len(hosts) == 0 {
 		return 0, fmt.Errorf("arg(hosts) empty")

alert/sender/lark.go

@@ -27,29 +27,29 @@ func (lk *LarkSender) CallBack(ctx CallBackContext) {
 		},
 	}
 
-	doSend(ctx.CallBackURL, body, models.Lark, ctx.Stats)
-	ctx.Stats.AlertNotifyTotal.WithLabelValues("rule_callback").Inc()
+	doSendAndRecord(ctx.Ctx, ctx.CallBackURL, ctx.CallBackURL, body, "callback", ctx.Stats, ctx.Events)
 }
 
 func (lk *LarkSender) Send(ctx MessageContext) {
 	if len(ctx.Users) == 0 || len(ctx.Events) == 0 {
 		return
 	}
-	urls := lk.extract(ctx.Users)
+	urls, tokens := lk.extract(ctx.Users)
 	message := BuildTplMessage(models.Lark, lk.tpl, ctx.Events)
-	for _, url := range urls {
+	for i, url := range urls {
 		body := feishu{
 			Msgtype: "text",
 			Content: feishuContent{
 				Text: message,
 			},
 		}
-		doSend(url, body, models.Lark, ctx.Stats)
+		doSendAndRecord(ctx.Ctx, url, tokens[i], body, models.Lark, ctx.Stats, ctx.Events)
 	}
 }
 
-func (lk *LarkSender) extract(users []*models.User) []string {
+func (lk *LarkSender) extract(users []*models.User) ([]string, []string) {
 	urls := make([]string, 0, len(users))
+	tokens := make([]string, 0, len(users))
 
 	for _, user := range users {
 		if token, has := user.ExtractToken(models.Lark); has {
@@ -58,7 +58,8 @@ func (lk *LarkSender) extract(users []*models.User) []string {
 				url = "https://open.larksuite.com/open-apis/bot/v2/hook/" + token
 			}
 			urls = append(urls, url)
+			tokens = append(tokens, token)
 		}
 	}
-	return urls
+	return urls, tokens
 }

alert/sender/larkcard.go

@@ -42,6 +42,7 @@ func (fs *LarkCardSender) CallBack(ctx CallBackContext) {
 	}
 
 	SendTitle := fmt.Sprintf("🔔 %s", ctx.Events[0].RuleName)
+	body := createFeishuCardBody()
 	body.Card.Header.Title.Content = SendTitle
 	body.Card.Header.Template = color
 	body.Card.Elements[0].Text.Content = message
@@ -55,14 +56,14 @@ func (fs *LarkCardSender) CallBack(ctx CallBackContext) {
 	}
 	parsedURL.RawQuery = ""
 
-	doSend(parsedURL.String(), body, models.LarkCard, ctx.Stats)
+	doSendAndRecord(ctx.Ctx, ctx.CallBackURL, ctx.CallBackURL, body, "callback", ctx.Stats, ctx.Events)
 }
 
 func (fs *LarkCardSender) Send(ctx MessageContext) {
 	if len(ctx.Users) == 0 || len(ctx.Events) == 0 {
 		return
 	}
-	urls, _ := fs.extract(ctx.Users)
+	urls, tokens := fs.extract(ctx.Users)
 	message := BuildTplMessage(models.LarkCard, fs.tpl, ctx.Events)
 	color := "red"
 	lowerUnicode := strings.ToLower(message)
@@ -73,18 +74,19 @@ func (fs *LarkCardSender) Send(ctx MessageContext) {
 	}
 
 	SendTitle := fmt.Sprintf("🔔 %s", ctx.Events[0].RuleName)
+	body := createFeishuCardBody()
 	body.Card.Header.Title.Content = SendTitle
 	body.Card.Header.Template = color
 	body.Card.Elements[0].Text.Content = message
 	body.Card.Elements[2].Elements[0].Content = SendTitle
-	for _, url := range urls {
-		doSend(url, body, models.LarkCard, ctx.Stats)
+	for i, url := range urls {
+		doSendAndRecord(ctx.Ctx, url, tokens[i], body, models.LarkCard, ctx.Stats, ctx.Events)
 	}
 }
 
 func (fs *LarkCardSender) extract(users []*models.User) ([]string, []string) {
 	urls := make([]string, 0, len(users))
-	ats := make([]string, 0)
+	tokens := make([]string, 0)
 	for i := range users {
 		if token, has := users[i].ExtractToken(models.Lark); has {
 			url := token
@@ -92,7 +94,8 @@ func (fs *LarkCardSender) extract(users []*models.User) ([]string, []string) {
 				url = "https://open.larksuite.com/open-apis/bot/v2/hook/" + strings.TrimSpace(token)
 			}
 			urls = append(urls, url)
+			tokens = append(tokens, token)
 		}
 	}
-	return urls, ats
+	return urls, tokens
 }

alert/sender/mm.go

@@ -7,6 +7,7 @@ import (
 
 	"github.com/ccfos/nightingale/v6/alert/astats"
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
 
 	"github.com/toolkits/pkg/logger"
 )
@@ -38,11 +39,11 @@ func (ms *MmSender) Send(ctx MessageContext) {
 	}
 	message := BuildTplMessage(models.Mm, ms.tpl, ctx.Events)
 
-	SendMM(MatterMostMessage{
+	SendMM(ctx.Ctx, MatterMostMessage{
 		Text:   message,
 		Tokens: urls,
 		Stats:  ctx.Stats,
-	})
+	}, ctx.Events, models.Mm)
 }
 
 func (ms *MmSender) CallBack(ctx CallBackContext) {
@@ -51,13 +52,11 @@ func (ms *MmSender) CallBack(ctx CallBackContext) {
 	}
 	message := BuildTplMessage(models.Mm, ms.tpl, ctx.Events)
 
-	SendMM(MatterMostMessage{
+	SendMM(ctx.Ctx, MatterMostMessage{
 		Text:   message,
 		Tokens: []string{ctx.CallBackURL},
 		Stats:  ctx.Stats,
-	})
-
-	ctx.Stats.AlertNotifyTotal.WithLabelValues("rule_callback").Inc()
+	}, ctx.Events, "callback")
 }
 
 func (ms *MmSender) extract(users []*models.User) []string {
@@ -70,11 +69,12 @@ func (ms *MmSender) extract(users []*models.User) []string {
 	return tokens
 }
 
-func SendMM(message MatterMostMessage) {
+func SendMM(ctx *ctx.Context, message MatterMostMessage, events []*models.AlertCurEvent, channel string) {
 	for i := 0; i < len(message.Tokens); i++ {
 		u, err := url.Parse(message.Tokens[i])
 		if err != nil {
 			logger.Errorf("mm_sender: failed to parse error=%v", err)
+			NotifyRecord(ctx, events, channel, message.Tokens[i], "", err)
 			continue
 		}
 
@@ -103,7 +103,7 @@ func SendMM(message MatterMostMessage) {
 				Username: username,
 				Text:     txt + message.Text,
 			}
-			doSend(ur, body, models.Mm, message.Stats)
+			doSendAndRecord(ctx, ur, message.Tokens[i], body, channel, message.Stats, events)
 		}
 	}
 }

alert/sender/plugin.go

@@ -2,26 +2,30 @@ package sender
 
 import (
 	"bytes"
+	"fmt"
 	"os"
 	"os/exec"
 	"time"
 
 	"github.com/ccfos/nightingale/v6/alert/astats"
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
 
 	"github.com/toolkits/pkg/file"
 	"github.com/toolkits/pkg/logger"
 	"github.com/toolkits/pkg/sys"
 )
 
-func MayPluginNotify(noticeBytes []byte, notifyScript models.NotifyScript, stats *astats.Stats) {
+func MayPluginNotify(ctx *ctx.Context, noticeBytes []byte, notifyScript models.NotifyScript,
+	stats *astats.Stats, event *models.AlertCurEvent) {
 	if len(noticeBytes) == 0 {
 		return
 	}
-	alertingCallScript(noticeBytes, notifyScript, stats)
+	alertingCallScript(ctx, noticeBytes, notifyScript, stats, event)
 }
 
-func alertingCallScript(stdinBytes []byte, notifyScript models.NotifyScript, stats *astats.Stats) {
+func alertingCallScript(ctx *ctx.Context, stdinBytes []byte, notifyScript models.NotifyScript,
+	stats *astats.Stats, event *models.AlertCurEvent) {
 	// not enable or no notify.py? do nothing
 	config := notifyScript
 	if !config.Enable || config.Content == "" {
@@ -81,6 +85,7 @@ func alertingCallScript(stdinBytes []byte, notifyScript models.NotifyScript, sta
 	}
 
 	err, isTimeout := sys.WrapTimeout(cmd, time.Duration(config.Timeout)*time.Second)
+	NotifyRecord(ctx, []*models.AlertCurEvent{event}, channel, cmd.String(), "", buildErr(err, isTimeout))
 
 	if isTimeout {
 		if err == nil {
@@ -102,3 +107,11 @@ func alertingCallScript(stdinBytes []byte, notifyScript models.NotifyScript, sta
 
 	logger.Infof("event_script_notify_ok: exec %s output: %s", fpath, buf.String())
 }
+
+func buildErr(err error, isTimeout bool) error {
+	if err == nil && !isTimeout {
+		return nil
+	} else {
+		return fmt.Errorf("is_timeout: %v, err: %v", isTimeout, err)
+	}
+}

alert/sender/sender.go

@@ -8,6 +8,7 @@ import (
 	"github.com/ccfos/nightingale/v6/alert/astats"
 	"github.com/ccfos/nightingale/v6/memsto"
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
 )
 
 type (
@@ -22,6 +23,7 @@ type (
 		Rule   *models.AlertRule
 		Events []*models.AlertCurEvent
 		Stats  *astats.Stats
+		Ctx    *ctx.Context
 	}
 )
 
@@ -49,13 +51,15 @@ func NewSender(key string, tpls map[string]*template.Template, smtp ...aconf.SMT
 	return nil
 }
 
-func BuildMessageContext(rule *models.AlertRule, events []*models.AlertCurEvent, uids []int64, userCache *memsto.UserCacheType, stats *astats.Stats) MessageContext {
+func BuildMessageContext(ctx *ctx.Context, rule *models.AlertRule, events []*models.AlertCurEvent,
+	uids []int64, userCache *memsto.UserCacheType, stats *astats.Stats) MessageContext {
 	users := userCache.GetByUserIds(uids)
 	return MessageContext{
 		Rule:   rule,
 		Events: events,
 		Users:  users,
 		Stats:  stats,
+		Ctx:    ctx,
 	}
 }
 

alert/sender/telegram.go

@@ -1,11 +1,13 @@
 package sender
 
 import (
+	"errors"
 	"html/template"
 	"strings"
 
 	"github.com/ccfos/nightingale/v6/alert/astats"
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
 
 	"github.com/toolkits/pkg/logger"
 )
@@ -35,13 +37,11 @@ func (ts *TelegramSender) CallBack(ctx CallBackContext) {
 	}
 
 	message := BuildTplMessage(models.Telegram, ts.tpl, ctx.Events)
-	SendTelegram(TelegramMessage{
+	SendTelegram(ctx.Ctx, TelegramMessage{
 		Text:   message,
 		Tokens: []string{ctx.CallBackURL},
 		Stats:  ctx.Stats,
-	})
-
-	ctx.Stats.AlertNotifyTotal.WithLabelValues("rule_callback").Inc()
+	}, ctx.Events, "callback")
 }
 
 func (ts *TelegramSender) Send(ctx MessageContext) {
@@ -51,11 +51,11 @@ func (ts *TelegramSender) Send(ctx MessageContext) {
 	tokens := ts.extract(ctx.Users)
 	message := BuildTplMessage(models.Telegram, ts.tpl, ctx.Events)
 
-	SendTelegram(TelegramMessage{
+	SendTelegram(ctx.Ctx, TelegramMessage{
 		Text:   message,
 		Tokens: tokens,
 		Stats:  ctx.Stats,
-	})
+	}, ctx.Events, models.Telegram)
 }
 
 func (ts *TelegramSender) extract(users []*models.User) []string {
@@ -68,10 +68,11 @@ func (ts *TelegramSender) extract(users []*models.User) []string {
 	return tokens
 }
 
-func SendTelegram(message TelegramMessage) {
+func SendTelegram(ctx *ctx.Context, message TelegramMessage, events []*models.AlertCurEvent, channel string) {
 	for i := 0; i < len(message.Tokens); i++ {
 		if !strings.Contains(message.Tokens[i], "/") && !strings.HasPrefix(message.Tokens[i], "https://") {
 			logger.Errorf("telegram_sender: result=fail invalid token=%s", message.Tokens[i])
+			NotifyRecord(ctx, events, channel, message.Tokens[i], "", errors.New("invalid token"))
 			continue
 		}
 		var url string
@@ -92,6 +93,6 @@ func SendTelegram(message TelegramMessage) {
 			Text:      message.Text,
 		}
 
-		doSend(url, body, models.Telegram, message.Stats)
+		doSendAndRecord(ctx, url, message.Tokens[i], body, channel, message.Stats, events)
 	}
 }

alert/sender/webhook.go

@@ -4,33 +4,41 @@ import (
 	"bytes"
 	"crypto/tls"
 	"encoding/json"
+	"fmt"
 	"io"
 	"net/http"
+	"sync"
 	"time"
 
 	"github.com/ccfos/nightingale/v6/alert/astats"
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
 
 	"github.com/toolkits/pkg/logger"
 )
 
-func sendWebhook(webhook *models.Webhook, event *models.AlertCurEvent, stats *astats.Stats) bool {
+func sendWebhook(webhook *models.Webhook, event interface{}, stats *astats.Stats) (bool, string, error) {
+	channel := "webhook"
+	if webhook.Type == models.RuleCallback {
+		channel = "callback"
+	}
+
 	conf := webhook
 	if conf.Url == "" || !conf.Enable {
-		return false
+		return false, "", nil
 	}
 	bs, err := json.Marshal(event)
 	if err != nil {
-		logger.Errorf("alertingWebhook failed to marshal event:%+v err:%v", event, err)
-		return false
+		logger.Errorf("%s alertingWebhook failed to marshal event:%+v err:%v", channel, event, err)
+		return false, "", err
 	}
 
 	bf := bytes.NewBuffer(bs)
 
 	req, err := http.NewRequest("POST", conf.Url, bf)
 	if err != nil {
-		logger.Warningf("alertingWebhook failed to new reques event:%+v err:%v", event, err)
-		return true
+		logger.Warningf("%s alertingWebhook failed to new reques event:%s err:%v", channel, string(bs), err)
+		return true, "", err
 	}
 
 	req.Header.Set("Content-Type", "application/json")
@@ -51,42 +59,48 @@ func sendWebhook(webhook *models.Webhook, event *models.AlertCurEvent, stats *as
 	if webhook != nil {
 		insecureSkipVerify = webhook.SkipVerify
 	}
-	client := http.Client{
-		Timeout: time.Duration(conf.Timeout) * time.Second,
-		Transport: &http.Transport{
-			TLSClientConfig: &tls.Config{InsecureSkipVerify: insecureSkipVerify},
-		},
+
+	if conf.Client == nil {
+		logger.Warningf("event_%s, event:%s, url: [%s], error: [%s]", channel, string(bs), conf.Url, "client is nil")
+		conf.Client = &http.Client{
+			Timeout: time.Duration(conf.Timeout) * time.Second,
+			Transport: &http.Transport{
+				TLSClientConfig: &tls.Config{InsecureSkipVerify: insecureSkipVerify},
+			},
+		}
 	}
 
-	stats.AlertNotifyTotal.WithLabelValues("webhook").Inc()
+	stats.AlertNotifyTotal.WithLabelValues(channel).Inc()
 	var resp *http.Response
-	resp, err = client.Do(req)
+	var body []byte
+	resp, err = conf.Client.Do(req)
+
 	if err != nil {
-		stats.AlertNotifyErrorTotal.WithLabelValues("webhook").Inc()
-		logger.Errorf("event_webhook_fail, ruleId: [%d], eventId: [%d], event:%+v, url: [%s], error: [%s]", event.RuleId, event.Id, event, conf.Url, err)
-		return true
+		stats.AlertNotifyErrorTotal.WithLabelValues(channel).Inc()
+		logger.Errorf("event_%s_fail, event:%s, url: [%s], error: [%s]", channel, string(bs), conf.Url, err)
+		return true, "", err
 	}
 
-	var body []byte
 	if resp.Body != nil {
 		defer resp.Body.Close()
 		body, _ = io.ReadAll(resp.Body)
 	}
 
 	if resp.StatusCode == 429 {
-		logger.Errorf("event_webhook_fail, url: %s, response code: %d, body: %s event:%+v", conf.Url, resp.StatusCode, string(body), event)
-		return true
+		logger.Errorf("event_%s_fail, url: %s, response code: %d, body: %s event:%s", channel, conf.Url, resp.StatusCode, string(body), string(bs))
+		return true, string(body), fmt.Errorf("status code is 429")
 	}
 
-	logger.Debugf("event_webhook_succ, url: %s, response code: %d, body: %s event:%+v", conf.Url, resp.StatusCode, string(body), event)
-	return false
+	logger.Debugf("event_%s_succ, url: %s, response code: %d, body: %s event:%s", channel, conf.Url, resp.StatusCode, string(body), string(bs))
+	return false, string(body), nil
 }
 
-func SendWebhooks(webhooks []*models.Webhook, event *models.AlertCurEvent, stats *astats.Stats) {
+func SingleSendWebhooks(ctx *ctx.Context, webhooks map[string]*models.Webhook, event *models.AlertCurEvent, stats *astats.Stats) {
 	for _, conf := range webhooks {
 		retryCount := 0
 		for retryCount < 3 {
-			needRetry := sendWebhook(conf, event, stats)
+			needRetry, res, err := sendWebhook(conf, event, stats)
+			NotifyRecord(ctx, []*models.AlertCurEvent{event}, "webhook", conf.Url, res, err)
 			if !needRetry {
 				break
 			}
@@ -95,3 +109,74 @@ func SendWebhooks(webhooks []*models.Webhook, event *models.AlertCurEvent, stats
 		}
 	}
 }
+
+func BatchSendWebhooks(ctx *ctx.Context, webhooks map[string]*models.Webhook, event *models.AlertCurEvent, stats *astats.Stats) {
+	for _, conf := range webhooks {
+		logger.Infof("push event:%+v to queue:%v", event, conf)
+		PushEvent(ctx, conf, event, stats)
+	}
+}
+
+var EventQueue = make(map[string]*WebhookQueue)
+var CallbackEventQueue = make(map[string]*WebhookQueue)
+var CallbackEventQueueLock sync.RWMutex
+var EventQueueLock sync.RWMutex
+
+const QueueMaxSize = 100000
+
+type WebhookQueue struct {
+	eventQueue *SafeEventQueue
+	closeCh    chan struct{}
+}
+
+func PushEvent(ctx *ctx.Context, webhook *models.Webhook, event *models.AlertCurEvent, stats *astats.Stats) {
+	EventQueueLock.RLock()
+	queue := EventQueue[webhook.Url]
+	EventQueueLock.RUnlock()
+
+	if queue == nil {
+		queue = &WebhookQueue{
+			eventQueue: NewSafeEventQueue(QueueMaxSize),
+			closeCh:    make(chan struct{}),
+		}
+
+		EventQueueLock.Lock()
+		EventQueue[webhook.Url] = queue
+		EventQueueLock.Unlock()
+
+		StartConsumer(ctx, queue, webhook.Batch, webhook, stats)
+	}
+
+	succ := queue.eventQueue.Push(event)
+	if !succ {
+		stats.AlertNotifyErrorTotal.WithLabelValues("push_event_queue").Inc()
+		logger.Warningf("Write channel(%s) full, current channel size: %d event:%v", webhook.Url, queue.eventQueue.Len(), event)
+	}
+}
+
+func StartConsumer(ctx *ctx.Context, queue *WebhookQueue, popSize int, webhook *models.Webhook, stats *astats.Stats) {
+	for {
+		select {
+		case <-queue.closeCh:
+			logger.Infof("event queue:%v closed", queue)
+			return
+		default:
+			events := queue.eventQueue.PopN(popSize)
+			if len(events) == 0 {
+				time.Sleep(time.Millisecond * 400)
+				continue
+			}
+
+			retryCount := 0
+			for retryCount < webhook.RetryCount {
+				needRetry, res, err := sendWebhook(webhook, events, stats)
+				go NotifyRecord(ctx, events, "webhook", webhook.Url, res, err)
+				if !needRetry {
+					break
+				}
+				retryCount++
+				time.Sleep(time.Second * time.Duration(webhook.RetryInterval) * time.Duration(retryCount))
+			}
+		}
+	}
+}

alert/sender/webhook_event_queue.go

@@ -0,0 +1,109 @@
+package sender
+
+import (
+	"container/list"
+	"sync"
+
+	"github.com/ccfos/nightingale/v6/models"
+)
+
+type SafeEventQueue struct {
+	lock        sync.RWMutex
+	maxSize     int
+	queueHigh   *list.List
+	queueMiddle *list.List
+	queueLow    *list.List
+}
+
+const (
+	High   = 1
+	Middle = 2
+	Low    = 3
+)
+
+func NewSafeEventQueue(maxSize int) *SafeEventQueue {
+	return &SafeEventQueue{
+		maxSize:     maxSize,
+		lock:        sync.RWMutex{},
+		queueHigh:   list.New(),
+		queueMiddle: list.New(),
+		queueLow:    list.New(),
+	}
+}
+
+func (spq *SafeEventQueue) Len() int {
+	spq.lock.RLock()
+	defer spq.lock.RUnlock()
+	return spq.queueHigh.Len() + spq.queueMiddle.Len() + spq.queueLow.Len()
+}
+
+// len 无锁读取长度，不要在本文件外调用
+func (spq *SafeEventQueue) len() int {
+	return spq.queueHigh.Len() + spq.queueMiddle.Len() + spq.queueLow.Len()
+}
+
+func (spq *SafeEventQueue) Push(event *models.AlertCurEvent) bool {
+	spq.lock.Lock()
+	defer spq.lock.Unlock()
+
+	for spq.len() > spq.maxSize {
+		return false
+	}
+
+	switch event.Severity {
+	case High:
+		spq.queueHigh.PushBack(event)
+	case Middle:
+		spq.queueMiddle.PushBack(event)
+	case Low:
+		spq.queueLow.PushBack(event)
+	default:
+		return false
+	}
+
+	return true
+}
+
+// pop 无锁弹出事件，不要在本文件外调用
+func (spq *SafeEventQueue) pop() *models.AlertCurEvent {
+	if spq.len() == 0 {
+		return nil
+	}
+
+	var elem interface{}
+
+	if spq.queueHigh.Len() > 0 {
+		elem = spq.queueHigh.Remove(spq.queueHigh.Front())
+	} else if spq.queueMiddle.Len() > 0 {
+		elem = spq.queueMiddle.Remove(spq.queueMiddle.Front())
+	} else {
+		elem = spq.queueLow.Remove(spq.queueLow.Front())
+	}
+	event, ok := elem.(*models.AlertCurEvent)
+	if !ok {
+		return nil
+	}
+	return event
+}
+
+func (spq *SafeEventQueue) Pop() *models.AlertCurEvent {
+	spq.lock.Lock()
+	defer spq.lock.Unlock()
+	return spq.pop()
+}
+
+func (spq *SafeEventQueue) PopN(n int) []*models.AlertCurEvent {
+	spq.lock.Lock()
+	defer spq.lock.Unlock()
+
+	events := make([]*models.AlertCurEvent, 0, n)
+	count := 0
+	for count < n && spq.len() > 0 {
+		event := spq.pop()
+		if event != nil {
+			events = append(events, event)
+		}
+		count++
+	}
+	return events
+}

alert/sender/webhook_event_queue_test.go

@@ -0,0 +1,157 @@
+package sender
+
+import (
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestSafePriorityQueue_ConcurrentPushPop(t *testing.T) {
+	spq := NewSafeEventQueue(100000)
+
+	var wg sync.WaitGroup
+	numGoroutines := 100
+	numEvents := 1000
+
+	// 并发 Push
+	wg.Add(numGoroutines)
+	for i := 0; i < numGoroutines; i++ {
+		go func(goroutineID int) {
+			defer wg.Done()
+			for j := 0; j < numEvents; j++ {
+				event := &models.AlertCurEvent{
+					Severity:    goroutineID%3 + 1,
+					TriggerTime: time.Now().UnixNano(),
+				}
+				spq.Push(event)
+			}
+		}(i)
+	}
+	wg.Wait()
+
+	// 检查队列长度是否正确
+	expectedLen := numGoroutines * numEvents
+	assert.Equal(t, expectedLen, spq.Len(), "Queue length mismatch after concurrent pushes")
+
+	// 并发 Pop
+	wg.Add(numGoroutines)
+	for i := 0; i < numGoroutines; i++ {
+		go func() {
+			defer wg.Done()
+			for {
+				event := spq.Pop()
+				if event == nil {
+					return
+				}
+			}
+		}()
+	}
+	wg.Wait()
+
+	// 最终队列应该为空
+	assert.Equal(t, 0, spq.Len(), "Queue should be empty after concurrent pops")
+}
+
+func TestSafePriorityQueue_ConcurrentPopMax(t *testing.T) {
+	spq := NewSafeEventQueue(100000)
+
+	// 添加初始数据
+	for i := 0; i < 1000; i++ {
+		spq.Push(&models.AlertCurEvent{
+			Severity:    i%3 + 1,
+			TriggerTime: time.Now().UnixNano(),
+		})
+	}
+
+	var wg sync.WaitGroup
+	numGoroutines := 10
+	popMax := 100
+
+	// 并发 PopN
+	wg.Add(numGoroutines)
+	for i := 0; i < numGoroutines; i++ {
+		go func() {
+			defer wg.Done()
+			events := spq.PopN(popMax)
+			assert.LessOrEqual(t, len(events), popMax, "PopN exceeded maximum")
+		}()
+	}
+	wg.Wait()
+
+	// 检查队列长度是否正确
+	expectedRemaining := 1000 - (numGoroutines * popMax)
+	if expectedRemaining < 0 {
+		expectedRemaining = 0
+	}
+	assert.Equal(t, expectedRemaining, spq.Len(), "Queue length mismatch after concurrent PopN")
+}
+
+func TestSafePriorityQueue_ConcurrentPushPopWithDifferentSeverities(t *testing.T) {
+	spq := NewSafeEventQueue(100000)
+
+	var wg sync.WaitGroup
+	numGoroutines := 50
+	numEvents := 500
+
+	// 并发 Push 不同优先级的事件
+	wg.Add(numGoroutines)
+	for i := 0; i < numGoroutines; i++ {
+		go func(goroutineID int) {
+			defer wg.Done()
+			for j := 0; j < numEvents; j++ {
+				event := &models.AlertCurEvent{
+					Severity:    goroutineID%3 + 1, // 模拟不同的 Severity
+					TriggerTime: time.Now().UnixNano(),
+				}
+				spq.Push(event)
+			}
+		}(i)
+	}
+	wg.Wait()
+
+	// 检查队列长度是否正确
+	expectedLen := numGoroutines * numEvents
+	assert.Equal(t, expectedLen, spq.Len(), "Queue length mismatch after concurrent pushes")
+
+	// 检查事件的顺序是否按照优先级排列
+	var lastEvent *models.AlertCurEvent
+	for spq.Len() > 0 {
+		event := spq.Pop()
+		if lastEvent != nil {
+			assert.LessOrEqual(t, lastEvent.Severity, event.Severity, "Events are not in correct priority order")
+		}
+		lastEvent = event
+	}
+}
+
+func TestSafePriorityQueue_ExceedMaxSize(t *testing.T) {
+	spq := NewSafeEventQueue(5)
+
+	// 插入超过最大容量的事件
+	for i := 0; i < 10; i++ {
+		spq.Push(&models.AlertCurEvent{
+			Severity:    i % 3,
+			TriggerTime: int64(i),
+		})
+	}
+
+	// 验证队列的长度是否不超过 maxSize
+	assert.LessOrEqual(t, spq.Len(), spq.maxSize)
+
+	// 验证队列中剩余事件的内容
+	expectedEvents := 5
+	if spq.Len() < 5 {
+		expectedEvents = spq.Len()
+	}
+
+	// 检查最后存入的事件是否是按优先级排序
+	for i := 0; i < expectedEvents; i++ {
+		event := spq.Pop()
+		if event != nil {
+			assert.LessOrEqual(t, event.Severity, 2)
+		}
+	}
+}

alert/sender/webhook_queue.go

@@ -0,0 +1,111 @@
+package sender
+
+import (
+	"container/list"
+	"sync"
+
+	"github.com/ccfos/nightingale/v6/models"
+)
+
+type SafeList struct {
+	sync.RWMutex
+	L *list.List
+}
+
+func NewSafeList() *SafeList {
+	return &SafeList{L: list.New()}
+}
+
+func (sl *SafeList) PushFront(v interface{}) *list.Element {
+	sl.Lock()
+	e := sl.L.PushFront(v)
+	sl.Unlock()
+	return e
+}
+
+func (sl *SafeList) PushFrontBatch(vs []interface{}) {
+	sl.Lock()
+	for _, item := range vs {
+		sl.L.PushFront(item)
+	}
+	sl.Unlock()
+}
+
+func (sl *SafeList) PopBack(max int) []*models.AlertCurEvent {
+	sl.Lock()
+
+	count := sl.L.Len()
+	if count == 0 {
+		sl.Unlock()
+		return []*models.AlertCurEvent{}
+	}
+
+	if count > max {
+		count = max
+	}
+
+	items := make([]*models.AlertCurEvent, 0, count)
+	for i := 0; i < count; i++ {
+		item := sl.L.Remove(sl.L.Back())
+		sample, ok := item.(*models.AlertCurEvent)
+		if ok {
+			items = append(items, sample)
+		}
+	}
+
+	sl.Unlock()
+	return items
+}
+
+func (sl *SafeList) RemoveAll() {
+	sl.Lock()
+	sl.L.Init()
+	sl.Unlock()
+}
+
+func (sl *SafeList) Len() int {
+	sl.RLock()
+	size := sl.L.Len()
+	sl.RUnlock()
+	return size
+}
+
+// SafeList with Limited Size
+type SafeListLimited struct {
+	maxSize int
+	SL      *SafeList
+}
+
+func NewSafeListLimited(maxSize int) *SafeListLimited {
+	return &SafeListLimited{SL: NewSafeList(), maxSize: maxSize}
+}
+
+func (sll *SafeListLimited) PopBack(max int) []*models.AlertCurEvent {
+	return sll.SL.PopBack(max)
+}
+
+func (sll *SafeListLimited) PushFront(v interface{}) bool {
+	if sll.SL.Len() >= sll.maxSize {
+		return false
+	}
+
+	sll.SL.PushFront(v)
+	return true
+}
+
+func (sll *SafeListLimited) PushFrontBatch(vs []interface{}) bool {
+	if sll.SL.Len() >= sll.maxSize {
+		return false
+	}
+
+	sll.SL.PushFrontBatch(vs)
+	return true
+}
+
+func (sll *SafeListLimited) RemoveAll() {
+	sll.SL.RemoveAll()
+}
+
+func (sll *SafeListLimited) Len() int {
+	return sll.SL.Len()
+}

alert/sender/wecom.go

@@ -37,29 +37,29 @@ func (ws *WecomSender) CallBack(ctx CallBackContext) {
 		},
 	}
 
-	doSend(ctx.CallBackURL, body, models.Wecom, ctx.Stats)
-	ctx.Stats.AlertNotifyTotal.WithLabelValues("rule_callback").Inc()
+	doSendAndRecord(ctx.Ctx, ctx.CallBackURL, ctx.CallBackURL, body, "callback", ctx.Stats, ctx.Events)
 }
 
 func (ws *WecomSender) Send(ctx MessageContext) {
 	if len(ctx.Users) == 0 || len(ctx.Events) == 0 {
 		return
 	}
-	urls := ws.extract(ctx.Users)
+	urls, tokens := ws.extract(ctx.Users)
 	message := BuildTplMessage(models.Wecom, ws.tpl, ctx.Events)
-	for _, url := range urls {
+	for i, url := range urls {
 		body := wecom{
 			Msgtype: "markdown",
 			Markdown: wecomMarkdown{
 				Content: message,
 			},
 		}
-		doSend(url, body, models.Wecom, ctx.Stats)
+		doSendAndRecord(ctx.Ctx, url, tokens[i], body, models.Wecom, ctx.Stats, ctx.Events)
 	}
 }
 
-func (ws *WecomSender) extract(users []*models.User) []string {
+func (ws *WecomSender) extract(users []*models.User) ([]string, []string) {
 	urls := make([]string, 0, len(users))
+	tokens := make([]string, 0, len(users))
 	for _, user := range users {
 		if token, has := user.ExtractToken(models.Wecom); has {
 			url := token
@@ -67,7 +67,8 @@ func (ws *WecomSender) extract(users []*models.User) []string {
 				url = "https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=" + token
 			}
 			urls = append(urls, url)
+			tokens = append(tokens, token)
 		}
 	}
-	return urls
+	return urls, tokens
 }

center/cconf/conf.go

@@ -13,6 +13,8 @@ type Center struct {
 	UseFileAssets          bool
 	FlashDuty              FlashDuty
 	EventHistoryGroupView  bool
+	CleanNotifyRecordDay   int
+	MigrateBusiGroupLabel  bool
 }
 
 type Plugin struct {

center/center.go

@@ -4,8 +4,11 @@ import (
 	"context"
 	"fmt"
 
+	"github.com/ccfos/nightingale/v6/dscache"
+
 	"github.com/ccfos/nightingale/v6/alert"
 	"github.com/ccfos/nightingale/v6/alert/astats"
+	"github.com/ccfos/nightingale/v6/alert/dispatch"
 	"github.com/ccfos/nightingale/v6/alert/process"
 	alertrt "github.com/ccfos/nightingale/v6/alert/router"
 	"github.com/ccfos/nightingale/v6/center/cconf"
@@ -16,6 +19,7 @@ import (
 	centerrt "github.com/ccfos/nightingale/v6/center/router"
 	"github.com/ccfos/nightingale/v6/center/sso"
 	"github.com/ccfos/nightingale/v6/conf"
+	"github.com/ccfos/nightingale/v6/cron"
 	"github.com/ccfos/nightingale/v6/dumper"
 	"github.com/ccfos/nightingale/v6/memsto"
 	"github.com/ccfos/nightingale/v6/models"
@@ -25,14 +29,13 @@ import (
 	"github.com/ccfos/nightingale/v6/pkg/httpx"
 	"github.com/ccfos/nightingale/v6/pkg/i18nx"
 	"github.com/ccfos/nightingale/v6/pkg/logx"
+	"github.com/ccfos/nightingale/v6/pkg/macros"
 	"github.com/ccfos/nightingale/v6/pkg/version"
 	"github.com/ccfos/nightingale/v6/prom"
 	"github.com/ccfos/nightingale/v6/pushgw/idents"
 	pushgwrt "github.com/ccfos/nightingale/v6/pushgw/router"
 	"github.com/ccfos/nightingale/v6/pushgw/writer"
 	"github.com/ccfos/nightingale/v6/storage"
-	"github.com/ccfos/nightingale/v6/tdengine"
-
 	"github.com/flashcatcloud/ibex/src/cmd/ibex"
 )
 
@@ -47,6 +50,10 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 
 	cconf.MergeOperationConf()
 
+	if config.Alert.Heartbeat.EngineName == "" {
+		config.Alert.Heartbeat.EngineName = "default"
+	}
+
 	logxClean, err := logx.Init(config.Log)
 	if err != nil {
 		return nil, err
@@ -62,7 +69,7 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 	}
 	ctx := ctx.NewContext(context.Background(), db, true)
 	migrate.Migrate(db)
-	models.InitRoot(ctx)
+	isRootInit := models.InitRoot(ctx)
 
 	config.HTTP.JWTAuth.SigningKey = models.InitJWTSigningKey(ctx)
 
@@ -71,7 +78,7 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 		return nil, err
 	}
 
-	integration.Init(ctx, config.Center.BuiltinIntegrationsDir)
+	go integration.Init(ctx, config.Center.BuiltinIntegrationsDir)
 	var redis storage.Redis
 	redis, err = storage.NewRedis(config.Redis)
 	if err != nil {
@@ -94,24 +101,38 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 	userCache := memsto.NewUserCache(ctx, syncStats)
 	userGroupCache := memsto.NewUserGroupCache(ctx, syncStats)
 	taskTplCache := memsto.NewTaskTplCache(ctx)
+	configCvalCache := memsto.NewCvalCache(ctx, syncStats)
 
 	sso := sso.Init(config.Center, ctx, configCache)
 	promClients := prom.NewPromClient(ctx)
-	tdengineClients := tdengine.NewTdengineClient(ctx, config.Alert.Heartbeat)
+
+	dispatch.InitRegisterQueryFunc(promClients)
 
 	externalProcessors := process.NewExternalProcessors()
-	alert.Start(config.Alert, config.Pushgw, syncStats, alertStats, externalProcessors, targetCache, busiGroupCache, alertMuteCache, alertRuleCache, notifyConfigCache, taskTplCache, dsCache, ctx, promClients, tdengineClients, userCache, userGroupCache)
+
+	macros.RegisterMacro(macros.MacroInVain)
+	dscache.Init(ctx, false)
+	alert.Start(config.Alert, config.Pushgw, syncStats, alertStats, externalProcessors, targetCache, busiGroupCache, alertMuteCache, alertRuleCache, notifyConfigCache, taskTplCache, dsCache, ctx, promClients, userCache, userGroupCache)
 
 	writers := writer.NewWriters(config.Pushgw)
 
 	go version.GetGithubVersion()
 
+	go cron.CleanNotifyRecord(ctx, config.Center.CleanNotifyRecordDay)
+
 	alertrtRouter := alertrt.New(config.HTTP, config.Alert, alertMuteCache, targetCache, busiGroupCache, alertStats, ctx, externalProcessors)
-	centerRouter := centerrt.New(config.HTTP, config.Center, config.Alert, config.Ibex, cconf.Operations, dsCache, notifyConfigCache, promClients, tdengineClients,
+	centerRouter := centerrt.New(config.HTTP, config.Center, config.Alert, config.Ibex,
+		cconf.Operations, dsCache, notifyConfigCache, promClients,
 		redis, sso, ctx, metas, idents, targetCache, userCache, userGroupCache)
 	pushgwRouter := pushgwrt.New(config.HTTP, config.Pushgw, config.Alert, targetCache, busiGroupCache, idents, metas, writers, ctx)
 
-	r := httpx.GinEngine(config.Global.RunMode, config.HTTP)
+	go func() {
+		if config.Center.MigrateBusiGroupLabel || models.CanMigrateBg(ctx) {
+			models.MigrateBg(ctx, pushgwRouter.Pushgw.BusiGroupLabelKey)
+		}
+	}()
+
+	r := httpx.GinEngine(config.Global.RunMode, config.HTTP, configCvalCache.PrintBodyPaths, configCvalCache.PrintAccessLog)
 
 	centerRouter.Config(r)
 	alertrtRouter.Config(r)
@@ -125,6 +146,11 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 
 	httpClean := httpx.Init(config.HTTP, r)
 
+	fmt.Printf("please view n9e at  http://%v:%v\n", config.Alert.Heartbeat.IP, config.HTTP.Port)
+	if isRootInit {
+		fmt.Println("username/password: root/root.2020")
+	}
+
 	return func() {
 		logxClean()
 		httpClean()

center/integration/init.go

@@ -16,6 +16,12 @@ import (
 const SYSTEM = "system"
 
 func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
+	err := models.InitBuiltinPayloads(ctx)
+	if err != nil {
+		logger.Warning("init old builtinPayloads fail ", err)
+		return
+	}
+
 	fp := builtinIntegrationsDir
 	if fp == "" {
 		fp = path.Join(runner.Cwd, "integrations")
@@ -92,6 +98,7 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 					logger.Warning("update builtin component fail ", old, err)
 				}
 			}
+			component.ID = old.ID
 		}
 
 		// delete uuid is emtpy
@@ -106,6 +113,12 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 			logger.Warning("delete builtin metrics fail ", err)
 		}
 
+		// 删除 uuid%1000 不为 0 uuid > 1000000000000000000 且 type 为 dashboard 的记录
+		err = models.DB(ctx).Exec("delete from builtin_payloads where uuid%1000 != 0 and uuid > 1000000000000000000 and type = 'dashboard' and updated_by = 'system'").Error
+		if err != nil {
+			logger.Warning("delete builtin payloads fail ", err)
+		}
+
 		// alerts
 		files, err = file.FilesUnder(componentDir + "/alerts")
 		if err == nil && len(files) > 0 {
@@ -141,13 +154,13 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 
 					cate := strings.Replace(f, ".json", "", -1)
 					builtinAlert := models.BuiltinPayload{
-						Component: component.Ident,
-						Type:      "alert",
-						Cate:      cate,
-						Name:      alert.Name,
-						Tags:      alert.AppendTags,
-						Content:   string(content),
-						UUID:      alert.UUID,
+						ComponentID: component.ID,
+						Type:        "alert",
+						Cate:        cate,
+						Name:        alert.Name,
+						Tags:        alert.AppendTags,
+						Content:     string(content),
+						UUID:        alert.UUID,
 					}
 
 					old, err := models.BuiltinPayloadGet(ctx, "uuid = ?", alert.UUID)
@@ -165,6 +178,7 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 					}
 
 					if old.UpdatedBy == SYSTEM {
+						old.ComponentID = component.ID
 						old.Content = string(content)
 						old.Name = alert.Name
 						old.Tags = alert.AppendTags
@@ -210,7 +224,8 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 				}
 
 				if dashboard.UUID == 0 {
-					dashboard.UUID = time.Now().UnixNano()
+					time.Sleep(time.Microsecond)
+					dashboard.UUID = time.Now().UnixMicro()
 					// 补全文件中的 uuid
 					bs, err = json.MarshalIndent(dashboard, "", "    ")
 					if err != nil {
@@ -231,13 +246,13 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 				}
 
 				builtinDashboard := models.BuiltinPayload{
-					Component: component.Ident,
-					Type:      "dashboard",
-					Cate:      "",
-					Name:      dashboard.Name,
-					Tags:      dashboard.Tags,
-					Content:   string(content),
-					UUID:      dashboard.UUID,
+					ComponentID: component.ID,
+					Type:        "dashboard",
+					Cate:        "",
+					Name:        dashboard.Name,
+					Tags:        dashboard.Tags,
+					Content:     string(content),
+					UUID:        dashboard.UUID,
 				}
 
 				old, err := models.BuiltinPayloadGet(ctx, "uuid = ?", dashboard.UUID)
@@ -255,6 +270,7 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 				}
 
 				if old.UpdatedBy == SYSTEM {
+					old.ComponentID = component.ID
 					old.Content = string(content)
 					old.Name = dashboard.Name
 					old.Tags = dashboard.Tags

center/metas/metas.go

@@ -92,7 +92,6 @@ func (s *Set) updateMeta(items map[string]models.HostMeta) {
 
 func (s *Set) updateTargets(m map[string]models.HostMeta) error {
 	if s.redis == nil {
-		logger.Warningf("redis is nil")
 		return nil
 	}
 

center/router/router.go

@@ -16,6 +16,7 @@ import (
 	"github.com/ccfos/nightingale/v6/conf"
 	_ "github.com/ccfos/nightingale/v6/front/statik"
 	"github.com/ccfos/nightingale/v6/memsto"
+	"github.com/ccfos/nightingale/v6/models"
 	"github.com/ccfos/nightingale/v6/pkg/aop"
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/pkg/httpx"
@@ -23,7 +24,6 @@ import (
 	"github.com/ccfos/nightingale/v6/prom"
 	"github.com/ccfos/nightingale/v6/pushgw/idents"
 	"github.com/ccfos/nightingale/v6/storage"
-	"github.com/ccfos/nightingale/v6/tdengine"
 
 	"github.com/gin-gonic/gin"
 	"github.com/rakyll/statik/fs"
@@ -41,7 +41,6 @@ type Router struct {
 	DatasourceCache   *memsto.DatasourceCacheType
 	NotifyConfigCache *memsto.NotifyConfigCacheType
 	PromClients       *prom.PromClientMap
-	TdendgineClients  *tdengine.TdengineClientMap
 	Redis             storage.Redis
 	MetaSet           *metas.Set
 	IdentSet          *idents.Set
@@ -51,9 +50,14 @@ type Router struct {
 	UserGroupCache    *memsto.UserGroupCacheType
 	Ctx               *ctx.Context
 	HeartbeatHook     HeartbeatHookFunc
+	TargetDeleteHook  models.TargetDeleteHookFunc
 }
 
-func New(httpConfig httpx.Config, center cconf.Center, alert aconf.Alert, ibex conf.Ibex, operations cconf.Operation, ds *memsto.DatasourceCacheType, ncc *memsto.NotifyConfigCacheType, pc *prom.PromClientMap, tdendgineClients *tdengine.TdengineClientMap, redis storage.Redis, sso *sso.SsoClient, ctx *ctx.Context, metaSet *metas.Set, idents *idents.Set, tc *memsto.TargetCacheType, uc *memsto.UserCacheType, ugc *memsto.UserGroupCacheType) *Router {
+func New(httpConfig httpx.Config, center cconf.Center, alert aconf.Alert, ibex conf.Ibex,
+	operations cconf.Operation, ds *memsto.DatasourceCacheType, ncc *memsto.NotifyConfigCacheType,
+	pc *prom.PromClientMap, redis storage.Redis,
+	sso *sso.SsoClient, ctx *ctx.Context, metaSet *metas.Set, idents *idents.Set,
+	tc *memsto.TargetCacheType, uc *memsto.UserCacheType, ugc *memsto.UserGroupCacheType) *Router {
 	return &Router{
 		HTTP:              httpConfig,
 		Center:            center,
@@ -63,7 +67,6 @@ func New(httpConfig httpx.Config, center cconf.Center, alert aconf.Alert, ibex c
 		DatasourceCache:   ds,
 		NotifyConfigCache: ncc,
 		PromClients:       pc,
-		TdendgineClients:  tdendgineClients,
 		Redis:             redis,
 		MetaSet:           metaSet,
 		IdentSet:          idents,
@@ -73,9 +76,14 @@ func New(httpConfig httpx.Config, center cconf.Center, alert aconf.Alert, ibex c
 		UserGroupCache:    ugc,
 		Ctx:               ctx,
 		HeartbeatHook:     func(ident string) map[string]interface{} { return nil },
+		TargetDeleteHook:  emptyDeleteHook,
 	}
 }
 
+func emptyDeleteHook(ctx *ctx.Context, idents []string) error {
+	return nil
+}
+
 func stat() gin.HandlerFunc {
 	return func(c *gin.Context) {
 		start := time.Now()
@@ -173,26 +181,53 @@ func (rt *Router) Config(r *gin.Engine) {
 			pages.POST("/query-range-batch", rt.promBatchQueryRange)
 			pages.POST("/query-instant-batch", rt.promBatchQueryInstant)
 			pages.GET("/datasource/brief", rt.datasourceBriefs)
+			pages.POST("/datasource/query", rt.datasourceQuery)
 
 			pages.POST("/ds-query", rt.QueryData)
-			pages.POST("/logs-query", rt.QueryLog)
+			pages.POST("/logs-query", rt.QueryLogV2)
 
 			pages.POST("/tdengine-databases", rt.tdengineDatabases)
 			pages.POST("/tdengine-tables", rt.tdengineTables)
 			pages.POST("/tdengine-columns", rt.tdengineColumns)
 
+			pages.POST("/log-query-batch", rt.QueryLogBatch)
+
+			// 数据库元数据接口
+			pages.POST("/db-databases", rt.ShowDatabases)
+			pages.POST("/db-tables", rt.ShowTables)
+			pages.POST("/db-desc-table", rt.DescribeTable)
+
+			// es 专用接口
+			pages.POST("/indices", rt.auth(), rt.user(), rt.QueryIndices)
+			pages.POST("/es-variable", rt.auth(), rt.user(), rt.QueryESVariable)
+			pages.POST("/fields", rt.auth(), rt.user(), rt.QueryFields)
+			pages.POST("/log-query", rt.auth(), rt.user(), rt.QueryLog)
 		} else {
 			pages.Any("/proxy/:id/*url", rt.auth(), rt.dsProxy)
 			pages.POST("/query-range-batch", rt.auth(), rt.promBatchQueryRange)
 			pages.POST("/query-instant-batch", rt.auth(), rt.promBatchQueryInstant)
 			pages.GET("/datasource/brief", rt.auth(), rt.user(), rt.datasourceBriefs)
+			pages.POST("/datasource/query", rt.auth(), rt.user(), rt.datasourceQuery)
 
 			pages.POST("/ds-query", rt.auth(), rt.QueryData)
-			pages.POST("/logs-query", rt.auth(), rt.QueryLog)
+			pages.POST("/logs-query", rt.auth(), rt.QueryLogV2)
 
 			pages.POST("/tdengine-databases", rt.auth(), rt.tdengineDatabases)
 			pages.POST("/tdengine-tables", rt.auth(), rt.tdengineTables)
 			pages.POST("/tdengine-columns", rt.auth(), rt.tdengineColumns)
+
+			pages.POST("/log-query-batch", rt.auth(), rt.user(), rt.QueryLogBatch)
+
+			// 数据库元数据接口
+			pages.POST("/db-databases", rt.auth(), rt.user(), rt.ShowDatabases)
+			pages.POST("/db-tables", rt.auth(), rt.user(), rt.ShowTables)
+			pages.POST("/db-desc-table", rt.auth(), rt.user(), rt.DescribeTable)
+
+			// es 专用接口
+			pages.POST("/indices", rt.auth(), rt.user(), rt.QueryIndices)
+			pages.POST("/es-variable", rt.QueryESVariable)
+			pages.POST("/fields", rt.QueryFields)
+			pages.POST("/log-query", rt.QueryLog)
 		}
 
 		pages.GET("/sql-template", rt.QuerySqlTemplate)
@@ -267,6 +302,7 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.DELETE("/busi-group/:id/members", rt.auth(), rt.user(), rt.perm("/busi-groups/put"), rt.bgrw(), rt.busiGroupMemberDel)
 		pages.DELETE("/busi-group/:id", rt.auth(), rt.user(), rt.perm("/busi-groups/del"), rt.bgrw(), rt.busiGroupDel)
 		pages.GET("/busi-group/:id/perm/:perm", rt.auth(), rt.user(), rt.checkBusiGroupPerm)
+		pages.GET("/busi-groups/tags", rt.auth(), rt.user(), rt.busiGroupsGetTags)
 
 		pages.GET("/targets", rt.auth(), rt.user(), rt.targetGets)
 		pages.GET("/target/extra-meta", rt.auth(), rt.user(), rt.targetExtendInfoByIdent)
@@ -276,7 +312,7 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.POST("/targets/tags", rt.auth(), rt.user(), rt.perm("/targets/put"), rt.targetBindTagsByFE)
 		pages.DELETE("/targets/tags", rt.auth(), rt.user(), rt.perm("/targets/put"), rt.targetUnbindTagsByFE)
 		pages.PUT("/targets/note", rt.auth(), rt.user(), rt.perm("/targets/put"), rt.targetUpdateNote)
-		pages.PUT("/targets/bgid", rt.auth(), rt.user(), rt.perm("/targets/put"), rt.targetUpdateBgid)
+		pages.PUT("/targets/bgids", rt.auth(), rt.user(), rt.perm("/targets/put"), rt.targetBindBgids)
 
 		pages.POST("/builtin-cate-favorite", rt.auth(), rt.user(), rt.builtinCateFavoriteAdd)
 		pages.DELETE("/builtin-cate-favorite/:name", rt.auth(), rt.user(), rt.builtinCateFavoriteDel)
@@ -297,6 +333,7 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.POST("/busi-group/:id/board/:bid/clone", rt.auth(), rt.user(), rt.perm("/dashboards/add"), rt.bgrw(), rt.boardClone)
 		pages.POST("/busi-groups/boards/clones", rt.auth(), rt.user(), rt.perm("/dashboards/add"), rt.boardBatchClone)
 
+		pages.GET("/boards", rt.auth(), rt.user(), rt.boardGetsByBids)
 		pages.GET("/board/:bid", rt.boardGet)
 		pages.GET("/board/:bid/pure", rt.boardPureGet)
 		pages.PUT("/board/:bid", rt.auth(), rt.user(), rt.perm("/dashboards/put"), rt.boardPut)
@@ -307,6 +344,11 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.GET("/share-charts", rt.chartShareGets)
 		pages.POST("/share-charts", rt.auth(), rt.chartShareAdd)
 
+		pages.POST("/dashboard-annotations", rt.auth(), rt.user(), rt.perm("/dashboards/put"), rt.dashAnnotationAdd)
+		pages.GET("/dashboard-annotations", rt.dashAnnotationGets)
+		pages.PUT("/dashboard-annotation/:id", rt.auth(), rt.user(), rt.perm("/dashboards/put"), rt.dashAnnotationPut)
+		pages.DELETE("/dashboard-annotation/:id", rt.auth(), rt.user(), rt.perm("/dashboards/del"), rt.dashAnnotationDel)
+
 		// pages.GET("/alert-rules/builtin/alerts-cates", rt.auth(), rt.user(), rt.builtinAlertCateGets)
 		// pages.GET("/alert-rules/builtin/list", rt.auth(), rt.user(), rt.builtinAlertRules)
 		pages.GET("/alert-rules/callbacks", rt.auth(), rt.user(), rt.alertRuleCallbacks)
@@ -315,6 +357,8 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.GET("/busi-group/:id/alert-rules", rt.auth(), rt.user(), rt.perm("/alert-rules"), rt.alertRuleGets)
 		pages.POST("/busi-group/:id/alert-rules", rt.auth(), rt.user(), rt.perm("/alert-rules/add"), rt.bgrw(), rt.alertRuleAddByFE)
 		pages.POST("/busi-group/:id/alert-rules/import", rt.auth(), rt.user(), rt.perm("/alert-rules/add"), rt.bgrw(), rt.alertRuleAddByImport)
+		pages.POST("/busi-group/:id/alert-rules/import-prom-rule", rt.auth(),
+			rt.user(), rt.perm("/alert-rules/add"), rt.bgrw(), rt.alertRuleAddByImportPromRule)
 		pages.DELETE("/busi-group/:id/alert-rules", rt.auth(), rt.user(), rt.perm("/alert-rules/del"), rt.bgrw(), rt.alertRuleDel)
 		pages.PUT("/busi-group/:id/alert-rules/fields", rt.auth(), rt.user(), rt.perm("/alert-rules/put"), rt.bgrw(), rt.alertRulePutFields)
 		pages.PUT("/busi-group/:id/alert-rule/:arid", rt.auth(), rt.user(), rt.perm("/alert-rules/put"), rt.alertRulePutByFE)
@@ -322,6 +366,7 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.GET("/alert-rule/:arid/pure", rt.auth(), rt.user(), rt.perm("/alert-rules"), rt.alertRulePureGet)
 		pages.PUT("/busi-group/alert-rule/validate", rt.auth(), rt.user(), rt.perm("/alert-rules/put"), rt.alertRuleValidation)
 		pages.POST("/relabel-test", rt.auth(), rt.user(), rt.relabelTest)
+		pages.POST("/busi-group/:id/alert-rules/clone", rt.auth(), rt.user(), rt.perm("/alert-rules/add"), rt.bgrw(), rt.cloneToMachine)
 
 		pages.GET("/busi-groups/recording-rules", rt.auth(), rt.user(), rt.perm("/recording-rules"), rt.recordingRuleGetsByGids)
 		pages.GET("/busi-group/:id/recording-rules", rt.auth(), rt.user(), rt.perm("/recording-rules"), rt.recordingRuleGets)
@@ -350,9 +395,11 @@ func (rt *Router) Config(r *gin.Engine) {
 		if rt.Center.AnonymousAccess.AlertDetail {
 			pages.GET("/alert-cur-event/:eid", rt.alertCurEventGet)
 			pages.GET("/alert-his-event/:eid", rt.alertHisEventGet)
+			pages.GET("/event-notify-records/:eid", rt.notificationRecordList)
 		} else {
 			pages.GET("/alert-cur-event/:eid", rt.auth(), rt.user(), rt.alertCurEventGet)
 			pages.GET("/alert-his-event/:eid", rt.auth(), rt.user(), rt.alertHisEventGet)
+			pages.GET("/event-notify-records/:eid", rt.auth(), rt.user(), rt.notificationRecordList)
 		}
 
 		// card logic
@@ -458,6 +505,8 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.GET("/builtin-payload/:id", rt.auth(), rt.user(), rt.perm("/built-in-components"), rt.builtinPayloadGet)
 		pages.PUT("/builtin-payloads", rt.auth(), rt.user(), rt.perm("/built-in-components/put"), rt.builtinPayloadsPut)
 		pages.DELETE("/builtin-payloads", rt.auth(), rt.user(), rt.perm("/built-in-components/del"), rt.builtinPayloadsDel)
+		pages.GET("/builtin-payload", rt.auth(), rt.user(), rt.builtinPayloadsGetByUUIDOrID)
+
 	}
 
 	r.GET("/api/n9e/versions", func(c *gin.Context) {
@@ -500,6 +549,8 @@ func (rt *Router) Config(r *gin.Engine) {
 			service.PUT("/targets/note", rt.targetUpdateNoteByService)
 			service.PUT("/targets/bgid", rt.targetUpdateBgidByService)
 
+			service.POST("/targets-of-host-query", rt.targetsOfHostQuery)
+
 			service.POST("/alert-rules", rt.alertRuleAddByService)
 			service.POST("/alert-rule-add", rt.alertRuleAddOneByService)
 			service.DELETE("/alert-rules", rt.alertRuleDelByService)
@@ -534,6 +585,7 @@ func (rt *Router) Config(r *gin.Engine) {
 			service.GET("/config/:id", rt.configGet)
 			service.GET("/configs", rt.configsGet)
 			service.GET("/config", rt.configGetByKey)
+			service.GET("/all-configs", rt.configGetAll)
 			service.PUT("/configs", rt.configsPut)
 			service.POST("/configs", rt.configsPost)
 			service.DELETE("/configs", rt.configsDel)
@@ -551,6 +603,11 @@ func (rt *Router) Config(r *gin.Engine) {
 
 			service.GET("/targets-of-alert-rule", rt.targetsOfAlertRule)
 
+			service.POST("/notify-record", rt.notificationRecordAdd)
+
+			service.GET("/alert-cur-events-del-by-hash", rt.alertCurEventDelByHash)
+
+			service.POST("/center/heartbeat", rt.heartbeat)
 		}
 	}
 

center/router/router_alert_cur_event.go

@@ -65,7 +65,8 @@ func (rt *Router) alertCurEventsCard(c *gin.Context) {
 	ginx.Dangerous(err)
 
 	// 最多获取50000个，获取太多也没啥意义
-	list, err := models.AlertCurEventGets(rt.Ctx, prods, bgids, stime, etime, severity, dsIds, cates, query, 50000, 0)
+	list, err := models.AlertCurEventsGet(rt.Ctx, prods, bgids, stime, etime, severity, dsIds,
+		cates, 0, query, 50000, 0)
 	ginx.Dangerous(err)
 
 	cardmap := make(map[string]*AlertCard)
@@ -162,13 +163,17 @@ func (rt *Router) alertCurEventsList(c *gin.Context) {
 		cates = strings.Split(cate, ",")
 	}
 
+	ruleId := ginx.QueryInt64(c, "rid", 0)
+
 	bgids, err := GetBusinessGroupIds(c, rt.Ctx, rt.Center.EventHistoryGroupView)
 	ginx.Dangerous(err)
 
-	total, err := models.AlertCurEventTotal(rt.Ctx, prods, bgids, stime, etime, severity, dsIds, cates, query)
+	total, err := models.AlertCurEventTotal(rt.Ctx, prods, bgids, stime, etime, severity, dsIds,
+		cates, ruleId, query)
 	ginx.Dangerous(err)
 
-	list, err := models.AlertCurEventGets(rt.Ctx, prods, bgids, stime, etime, severity, dsIds, cates, query, limit, ginx.Offset(c, limit))
+	list, err := models.AlertCurEventsGet(rt.Ctx, prods, bgids, stime, etime, severity, dsIds,
+		cates, ruleId, query, limit, ginx.Offset(c, limit))
 	ginx.Dangerous(err)
 
 	cache := make(map[int64]*models.UserGroup)
@@ -201,7 +206,9 @@ func (rt *Router) checkCurEventBusiGroupRWPermission(c *gin.Context, ids []int64
 	for i := 0; i < len(ids); i++ {
 		event, err := models.AlertCurEventGetById(rt.Ctx, ids[i])
 		ginx.Dangerous(err)
-
+		if event == nil {
+			continue
+		}
 		if _, has := set[event.GroupId]; !has {
 			rt.bgrwCheck(c, event.GroupId)
 			set[event.GroupId] = struct{}{}
@@ -222,6 +229,12 @@ func (rt *Router) alertCurEventGet(c *gin.Context) {
 		rt.bgroCheck(c, event.GroupId)
 	}
 
+	ruleConfig, needReset := models.FillRuleConfigTplName(rt.Ctx, event.RuleConfig)
+	if needReset {
+		event.RuleConfigJson = ruleConfig
+	}
+
+	event.LastEvalTime = event.TriggerTime
 	ginx.NewRender(c).Data(event, nil)
 }
 
@@ -229,3 +242,8 @@ func (rt *Router) alertCurEventsStatistics(c *gin.Context) {
 
 	ginx.NewRender(c).Data(models.AlertCurEventStatistics(rt.Ctx, time.Now()), nil)
 }
+
+func (rt *Router) alertCurEventDelByHash(c *gin.Context) {
+	hash := ginx.QueryStr(c, "hash")
+	ginx.NewRender(c).Message(models.AlertCurEventDelByHash(rt.Ctx, hash))
+}

center/router/router_alert_his_event.go

@@ -54,13 +54,17 @@ func (rt *Router) alertHisEventsList(c *gin.Context) {
 		cates = strings.Split(cate, ",")
 	}
 
+	ruleId := ginx.QueryInt64(c, "rid", 0)
+
 	bgids, err := GetBusinessGroupIds(c, rt.Ctx, rt.Center.EventHistoryGroupView)
 	ginx.Dangerous(err)
 
-	total, err := models.AlertHisEventTotal(rt.Ctx, prods, bgids, stime, etime, severity, recovered, dsIds, cates, query)
+	total, err := models.AlertHisEventTotal(rt.Ctx, prods, bgids, stime, etime, severity,
+		recovered, dsIds, cates, ruleId, query)
 	ginx.Dangerous(err)
 
-	list, err := models.AlertHisEventGets(rt.Ctx, prods, bgids, stime, etime, severity, recovered, dsIds, cates, query, limit, ginx.Offset(c, limit))
+	list, err := models.AlertHisEventGets(rt.Ctx, prods, bgids, stime, etime, severity, recovered,
+		dsIds, cates, ruleId, query, limit, ginx.Offset(c, limit))
 	ginx.Dangerous(err)
 
 	cache := make(map[int64]*models.UserGroup)
@@ -87,6 +91,11 @@ func (rt *Router) alertHisEventGet(c *gin.Context) {
 		rt.bgroCheck(c, event.GroupId)
 	}
 
+	ruleConfig, needReset := models.FillRuleConfigTplName(rt.Ctx, event.RuleConfig)
+	if needReset {
+		event.RuleConfigJson = ruleConfig
+	}
+
 	ginx.NewRender(c).Data(event, err)
 }
 

center/router/router_alert_rule.go

@@ -1,17 +1,22 @@
 package router
 
 import (
+	"encoding/json"
 	"fmt"
 	"net/http"
+	"regexp"
 	"strconv"
 	"strings"
 	"time"
 
+	"gopkg.in/yaml.v2"
+
 	"github.com/ccfos/nightingale/v6/models"
 	"github.com/ccfos/nightingale/v6/pushgw/pconf"
 	"github.com/ccfos/nightingale/v6/pushgw/writer"
 
 	"github.com/gin-gonic/gin"
+	"github.com/jinzhu/copier"
 	"github.com/prometheus/prometheus/prompb"
 	"github.com/toolkits/pkg/ginx"
 	"github.com/toolkits/pkg/i18n"
@@ -32,6 +37,18 @@ func (rt *Router) alertRuleGets(c *gin.Context) {
 	ginx.NewRender(c).Data(ars, err)
 }
 
+func getAlertCueEventTimeRange(c *gin.Context) (stime, etime int64) {
+	stime = ginx.QueryInt64(c, "stime", 0)
+	etime = ginx.QueryInt64(c, "etime", 0)
+	if etime == 0 {
+		etime = time.Now().Unix()
+	}
+	if stime == 0 || stime >= etime {
+		stime = etime - 30*24*int64(time.Hour.Seconds())
+	}
+	return
+}
+
 func (rt *Router) alertRuleGetsByGids(c *gin.Context) {
 	gids := str.IdsInt64(ginx.QueryStr(c, "gids", ""), ",")
 	if len(gids) > 0 {
@@ -55,9 +72,35 @@ func (rt *Router) alertRuleGetsByGids(c *gin.Context) {
 	ars, err := models.AlertRuleGetsByBGIds(rt.Ctx, gids)
 	if err == nil {
 		cache := make(map[int64]*models.UserGroup)
+		rids := make([]int64, 0, len(ars))
+		names := make([]string, 0, len(ars))
 		for i := 0; i < len(ars); i++ {
 			ars[i].FillNotifyGroups(rt.Ctx, cache)
 			ars[i].FillSeverities()
+
+			if len(ars[i].DatasourceQueries) != 0 {
+				ars[i].DatasourceIdsJson = rt.DatasourceCache.GetIDsByDsCateAndQueries(ars[i].Cate, ars[i].DatasourceQueries)
+			}
+
+			rids = append(rids, ars[i].Id)
+			names = append(names, ars[i].UpdateBy)
+		}
+
+		stime, etime := getAlertCueEventTimeRange(c)
+		cnt := models.AlertCurEventCountByRuleId(rt.Ctx, rids, stime, etime)
+		if cnt != nil {
+			for i := 0; i < len(ars); i++ {
+				ars[i].CurEventCount = cnt[ars[i].Id]
+			}
+		}
+
+		users := models.UserMapGet(rt.Ctx, "username in (?)", names)
+		if users != nil {
+			for i := 0; i < len(ars); i++ {
+				if user, exist := users[ars[i].UpdateBy]; exist {
+					ars[i].UpdateByNickname = user.Nickname
+				}
+			}
 		}
 	}
 	ginx.NewRender(c).Data(ars, err)
@@ -85,6 +128,10 @@ func (rt *Router) alertRulesGetByService(c *gin.Context) {
 		cache := make(map[int64]*models.UserGroup)
 		for i := 0; i < len(ars); i++ {
 			ars[i].FillNotifyGroups(rt.Ctx, cache)
+
+			if len(ars[i].DatasourceQueries) != 0 {
+				ars[i].DatasourceIdsJson = rt.DatasourceCache.GetIDsByDsCateAndQueries(ars[i].Cate, ars[i].DatasourceQueries)
+			}
 		}
 	}
 	ginx.NewRender(c).Data(ars, err)
@@ -119,12 +166,48 @@ func (rt *Router) alertRuleAddByImport(c *gin.Context) {
 		ginx.Bomb(http.StatusBadRequest, "input json is empty")
 	}
 
+	for i := range lst {
+		if len(lst[i].DatasourceQueries) == 0 {
+			lst[i].DatasourceQueries = []models.DatasourceQuery{
+				models.DataSourceQueryAll,
+			}
+		}
+	}
+
 	bgid := ginx.UrlParamInt64(c, "id")
 	reterr := rt.alertRuleAdd(lst, username, bgid, c.GetHeader("X-Language"))
 
 	ginx.NewRender(c).Data(reterr, nil)
 }
 
+type promRuleForm struct {
+	Payload           string                   `json:"payload" binding:"required"`
+	DatasourceQueries []models.DatasourceQuery `json:"datasource_queries" binding:"required"`
+	Disabled          int                      `json:"disabled" binding:"gte=0,lte=1"`
+}
+
+func (rt *Router) alertRuleAddByImportPromRule(c *gin.Context) {
+	var f promRuleForm
+	ginx.Dangerous(c.BindJSON(&f))
+
+	var pr struct {
+		Groups []models.PromRuleGroup `yaml:"groups"`
+	}
+	err := yaml.Unmarshal([]byte(f.Payload), &pr)
+	if err != nil {
+		ginx.Bomb(http.StatusBadRequest, "invalid yaml format, please use the example format. err: %v", err)
+	}
+
+	if len(pr.Groups) == 0 {
+		ginx.Bomb(http.StatusBadRequest, "input yaml is empty")
+	}
+
+	lst := models.DealPromGroup(pr.Groups, f.DatasourceQueries, f.Disabled)
+	username := c.MustGet("username").(string)
+	bgid := ginx.UrlParamInt64(c, "id")
+	ginx.NewRender(c).Data(rt.alertRuleAdd(lst, username, bgid, c.GetHeader("X-Language")), nil)
+}
+
 func (rt *Router) alertRuleAddByService(c *gin.Context) {
 	var lst []models.AlertRule
 	ginx.BindJSON(c, &lst)
@@ -275,6 +358,43 @@ func (rt *Router) alertRulePutFields(c *gin.Context) {
 			continue
 		}
 
+		if f.Action == "update_triggers" {
+			if triggers, has := f.Fields["triggers"]; has {
+				originRule := ar.RuleConfigJson.(map[string]interface{})
+				originRule["triggers"] = triggers
+				b, err := json.Marshal(originRule)
+				ginx.Dangerous(err)
+				ginx.Dangerous(ar.UpdateFieldsMap(rt.Ctx, map[string]interface{}{"rule_config": string(b)}))
+				continue
+			}
+		}
+
+		if f.Action == "annotations_add" {
+			if annotations, has := f.Fields["annotations"]; has {
+				annotationsMap := annotations.(map[string]interface{})
+				for k, v := range annotationsMap {
+					ar.AnnotationsJSON[k] = v.(string)
+				}
+				b, err := json.Marshal(ar.AnnotationsJSON)
+				ginx.Dangerous(err)
+				ginx.Dangerous(ar.UpdateFieldsMap(rt.Ctx, map[string]interface{}{"annotations": string(b)}))
+				continue
+			}
+		}
+
+		if f.Action == "annotations_del" {
+			if annotations, has := f.Fields["annotations"]; has {
+				annotationsKeys := annotations.(map[string]interface{})
+				for key := range annotationsKeys {
+					delete(ar.AnnotationsJSON, key)
+				}
+				b, err := json.Marshal(ar.AnnotationsJSON)
+				ginx.Dangerous(err)
+				ginx.Dangerous(ar.UpdateFieldsMap(rt.Ctx, map[string]interface{}{"annotations": string(b)}))
+				continue
+			}
+		}
+
 		if f.Action == "callback_add" {
 			// 增加一个 callback 地址
 			if callbacks, has := f.Fields["callbacks"]; has {
@@ -295,6 +415,16 @@ func (rt *Router) alertRulePutFields(c *gin.Context) {
 			}
 		}
 
+		if f.Action == "datasource_change" {
+			// 修改数据源
+			if datasourceQueries, has := f.Fields["datasource_queries"]; has {
+				bytes, err := json.Marshal(datasourceQueries)
+				ginx.Dangerous(err)
+				ginx.Dangerous(ar.UpdateFieldsMap(rt.Ctx, map[string]interface{}{"datasource_queries": bytes}))
+				continue
+			}
+		}
+
 		for k, v := range f.Fields {
 			ginx.Dangerous(ar.UpdateColumn(rt.Ctx, k, v))
 		}
@@ -314,6 +444,10 @@ func (rt *Router) alertRuleGet(c *gin.Context) {
 		return
 	}
 
+	if len(ar.DatasourceQueries) != 0 {
+		ar.DatasourceIdsJson = rt.DatasourceCache.GetIDsByDsCateAndQueries(ar.Cate, ar.DatasourceQueries)
+	}
+
 	err = ar.FillNotifyGroups(rt.Ctx, make(map[int64]*models.UserGroup))
 	ginx.Dangerous(err)
 
@@ -422,7 +556,7 @@ func (rt *Router) relabelTest(c *gin.Context) {
 
 	labels := make([]prompb.Label, len(f.Tags))
 	for i, tag := range f.Tags {
-		label := strings.Split(tag, "=")
+		label := strings.SplitN(tag, "=", 2)
 		if len(label) != 2 {
 			ginx.Bomb(http.StatusBadRequest, "tag:%s format error", tag)
 		}
@@ -453,3 +587,91 @@ func (rt *Router) relabelTest(c *gin.Context) {
 
 	ginx.NewRender(c).Data(tags, nil)
 }
+
+type identListForm struct {
+	Ids       []int64  `json:"ids"`
+	IdentList []string `json:"ident_list"`
+}
+
+func containsIdentOperator(s string) bool {
+	pattern := `ident\s*(!=|!~|=~)`
+	matched, err := regexp.MatchString(pattern, s)
+	if err != nil {
+		return false
+	}
+	return matched
+}
+
+func (rt *Router) cloneToMachine(c *gin.Context) {
+	var f identListForm
+	ginx.BindJSON(c, &f)
+
+	if len(f.IdentList) == 0 {
+		ginx.Bomb(http.StatusBadRequest, "ident_list is empty")
+	}
+
+	alertRules, err := models.AlertRuleGetsByIds(rt.Ctx, f.Ids)
+	ginx.Dangerous(err)
+
+	re := regexp.MustCompile(`ident\s*=\s*\\".*?\\"`)
+
+	user := c.MustGet("username").(string)
+	now := time.Now().Unix()
+
+	newRules := make([]*models.AlertRule, 0)
+
+	reterr := make(map[string]map[string]string)
+
+	for i := range alertRules {
+		errMsg := make(map[string]string)
+
+		if alertRules[i].Cate != "prometheus" {
+			errMsg["all"] = "Only Prometheus rule can be cloned to machines"
+			reterr[alertRules[i].Name] = errMsg
+			continue
+		}
+
+		if containsIdentOperator(alertRules[i].RuleConfig) {
+			errMsg["all"] = "promql is missing ident"
+			reterr[alertRules[i].Name] = errMsg
+			continue
+		}
+
+		for j := range f.IdentList {
+			alertRules[i].RuleConfig = re.ReplaceAllString(alertRules[i].RuleConfig, fmt.Sprintf(`ident=\"%s\"`, f.IdentList[j]))
+
+			newRule := &models.AlertRule{}
+			if err := copier.Copy(newRule, alertRules[i]); err != nil {
+				errMsg[f.IdentList[j]] = fmt.Sprintf("fail to clone rule, err: %s", err)
+				continue
+			}
+
+			newRule.Id = 0
+			newRule.Name = alertRules[i].Name + "_" + f.IdentList[j]
+			newRule.CreateBy = user
+			newRule.UpdateBy = user
+			newRule.UpdateAt = now
+			newRule.CreateAt = now
+			newRule.RuleConfig = alertRules[i].RuleConfig
+
+			exist, err := models.AlertRuleExists(rt.Ctx, 0, newRule.GroupId, newRule.Name)
+			if err != nil {
+				errMsg[f.IdentList[j]] = err.Error()
+				continue
+			}
+
+			if exist {
+				errMsg[f.IdentList[j]] = fmt.Sprintf("rule already exists, ruleName: %s", newRule.Name)
+				continue
+			}
+
+			newRules = append(newRules, newRule)
+		}
+
+		if len(errMsg) > 0 {
+			reterr[alertRules[i].Name] = errMsg
+		}
+	}
+
+	ginx.NewRender(c).Data(reterr, models.InsertAlertRule(rt.Ctx, newRules))
+}

center/router/router_board.go

@@ -94,6 +94,14 @@ func (rt *Router) boardGet(c *gin.Context) {
 	ginx.NewRender(c).Data(board, nil)
 }
 
+// 根据 bids 参数，获取多个 board
+func (rt *Router) boardGetsByBids(c *gin.Context) {
+	bids := str.IdsInt64(ginx.QueryStr(c, "bids", ""), ",")
+	boards, err := models.BoardGetsByBids(rt.Ctx, bids)
+	ginx.Dangerous(err)
+	ginx.NewRender(c).Data(boards, err)
+}
+
 func (rt *Router) boardPureGet(c *gin.Context) {
 	board, err := models.BoardGetByID(rt.Ctx, ginx.UrlParamInt64(c, "bid"))
 	ginx.Dangerous(err)

center/router/router_builtin_componet.go

@@ -4,10 +4,15 @@ import (
 	"net/http"
 
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
+
 	"github.com/gin-gonic/gin"
 	"github.com/toolkits/pkg/ginx"
+	"gorm.io/gorm"
 )
 
+const SYSTEM = "system"
+
 func (rt *Router) builtinComponentsAdd(c *gin.Context) {
 	var lst []models.BuiltinComponent
 	ginx.BindJSON(c, &lst)
@@ -31,8 +36,9 @@ func (rt *Router) builtinComponentsAdd(c *gin.Context) {
 
 func (rt *Router) builtinComponentsGets(c *gin.Context) {
 	query := ginx.QueryStr(c, "query", "")
+	disabled := ginx.QueryInt(c, "disabled", -1)
 
-	bc, err := models.BuiltinComponentGets(rt.Ctx, query)
+	bc, err := models.BuiltinComponentGets(rt.Ctx, query, disabled)
 	ginx.Dangerous(err)
 
 	ginx.NewRender(c).Data(bc, nil)
@@ -50,10 +56,31 @@ func (rt *Router) builtinComponentsPut(c *gin.Context) {
 		return
 	}
 
+	if bc.CreatedBy == SYSTEM {
+		req.Ident = bc.Ident
+	}
+
 	username := Username(c)
 	req.UpdatedBy = username
 
-	ginx.NewRender(c).Message(bc.Update(rt.Ctx, req))
+	err = models.DB(rt.Ctx).Transaction(func(tx *gorm.DB) error {
+		tCtx := &ctx.Context{
+			DB: tx,
+		}
+
+		txErr := models.BuiltinMetricBatchUpdateColumn(tCtx, "typ", bc.Ident, req.Ident, req.UpdatedBy)
+		if txErr != nil {
+			return txErr
+		}
+
+		txErr = bc.Update(tCtx, req)
+		if txErr != nil {
+			return txErr
+		}
+		return nil
+	})
+
+	ginx.NewRender(c).Message(err)
 }
 
 func (rt *Router) builtinComponentsDel(c *gin.Context) {

center/router/router_builtin_metrics.go

@@ -102,9 +102,29 @@ func (rt *Router) builtinMetricsDefaultTypes(c *gin.Context) {
 func (rt *Router) builtinMetricsTypes(c *gin.Context) {
 	collector := ginx.QueryStr(c, "collector", "")
 	query := ginx.QueryStr(c, "query", "")
+	disabled := ginx.QueryInt(c, "disabled", -1)
 	lang := c.GetHeader("X-Language")
 
-	ginx.NewRender(c).Data(models.BuiltinMetricTypes(rt.Ctx, lang, collector, query))
+	metricTypeList, err := models.BuiltinMetricTypes(rt.Ctx, lang, collector, query)
+	ginx.Dangerous(err)
+
+	componentList, err := models.BuiltinComponentGets(rt.Ctx, "", disabled)
+	ginx.Dangerous(err)
+
+	// 创建一个 map 来存储 componentList 中的类型
+	componentTypes := make(map[string]struct{})
+	for _, comp := range componentList {
+		componentTypes[comp.Ident] = struct{}{}
+	}
+
+	filteredMetricTypeList := make([]string, 0)
+	for _, metricType := range metricTypeList {
+		if _, exists := componentTypes[metricType]; exists {
+			filteredMetricTypeList = append(filteredMetricTypeList, metricType)
+		}
+	}
+
+	ginx.NewRender(c).Data(filteredMetricTypeList, nil)
 }
 
 func (rt *Router) builtinMetricsCollectors(c *gin.Context) {

center/router/router_builtin_payload.go

@@ -6,6 +6,7 @@ import (
 	"strings"
 	"time"
 
+	"github.com/BurntSushi/toml"
 	"github.com/ccfos/nightingale/v6/models"
 	"github.com/gin-gonic/gin"
 	"github.com/toolkits/pkg/ginx"
@@ -42,7 +43,7 @@ func (rt *Router) builtinPayloadsAdd(c *gin.Context) {
 
 				for _, rule := range alertRules {
 					if rule.UUID == 0 {
-						rule.UUID = time.Now().UnixNano()
+						rule.UUID = time.Now().UnixMicro()
 					}
 
 					contentBytes, err := json.Marshal(rule)
@@ -52,15 +53,15 @@ func (rt *Router) builtinPayloadsAdd(c *gin.Context) {
 					}
 
 					bp := models.BuiltinPayload{
-						Type:      lst[i].Type,
-						Component: lst[i].Component,
-						Cate:      lst[i].Cate,
-						Name:      rule.Name,
-						Tags:      rule.AppendTags,
-						UUID:      rule.UUID,
-						Content:   string(contentBytes),
-						CreatedBy: username,
-						UpdatedBy: username,
+						Type:        lst[i].Type,
+						ComponentID: lst[i].ComponentID,
+						Cate:        lst[i].Cate,
+						Name:        rule.Name,
+						Tags:        rule.AppendTags,
+						UUID:        rule.UUID,
+						Content:     string(contentBytes),
+						CreatedBy:   username,
+						UpdatedBy:   username,
 					}
 
 					if err := bp.Add(rt.Ctx, username); err != nil {
@@ -77,19 +78,25 @@ func (rt *Router) builtinPayloadsAdd(c *gin.Context) {
 			}
 
 			if alertRule.UUID == 0 {
-				alertRule.UUID = time.Now().UnixNano()
+				alertRule.UUID = time.Now().UnixMicro()
+			}
+
+			contentBytes, err := json.Marshal(alertRule)
+			if err != nil {
+				reterr[alertRule.Name] = err.Error()
+				continue
 			}
 
 			bp := models.BuiltinPayload{
-				Type:      lst[i].Type,
-				Component: lst[i].Component,
-				Cate:      lst[i].Cate,
-				Name:      alertRule.Name,
-				Tags:      alertRule.AppendTags,
-				UUID:      alertRule.UUID,
-				Content:   lst[i].Content,
-				CreatedBy: username,
-				UpdatedBy: username,
+				Type:        lst[i].Type,
+				ComponentID: lst[i].ComponentID,
+				Cate:        lst[i].Cate,
+				Name:        alertRule.Name,
+				Tags:        alertRule.AppendTags,
+				UUID:        alertRule.UUID,
+				Content:     string(contentBytes),
+				CreatedBy:   username,
+				UpdatedBy:   username,
 			}
 
 			if err := bp.Add(rt.Ctx, username); err != nil {
@@ -105,7 +112,7 @@ func (rt *Router) builtinPayloadsAdd(c *gin.Context) {
 
 				for _, dashboard := range dashboards {
 					if dashboard.UUID == 0 {
-						dashboard.UUID = time.Now().UnixNano()
+						dashboard.UUID = time.Now().UnixMicro()
 					}
 
 					contentBytes, err := json.Marshal(dashboard)
@@ -115,15 +122,15 @@ func (rt *Router) builtinPayloadsAdd(c *gin.Context) {
 					}
 
 					bp := models.BuiltinPayload{
-						Type:      lst[i].Type,
-						Component: lst[i].Component,
-						Cate:      lst[i].Cate,
-						Name:      dashboard.Name,
-						Tags:      dashboard.Tags,
-						UUID:      dashboard.UUID,
-						Content:   string(contentBytes),
-						CreatedBy: username,
-						UpdatedBy: username,
+						Type:        lst[i].Type,
+						ComponentID: lst[i].ComponentID,
+						Cate:        lst[i].Cate,
+						Name:        dashboard.Name,
+						Tags:        dashboard.Tags,
+						UUID:        dashboard.UUID,
+						Content:     string(contentBytes),
+						CreatedBy:   username,
+						UpdatedBy:   username,
 					}
 
 					if err := bp.Add(rt.Ctx, username); err != nil {
@@ -140,25 +147,39 @@ func (rt *Router) builtinPayloadsAdd(c *gin.Context) {
 			}
 
 			if dashboard.UUID == 0 {
-				dashboard.UUID = time.Now().UnixNano()
+				dashboard.UUID = time.Now().UnixMicro()
+			}
+
+			contentBytes, err := json.Marshal(dashboard)
+			if err != nil {
+				reterr[dashboard.Name] = err.Error()
+				continue
 			}
 
 			bp := models.BuiltinPayload{
-				Type:      lst[i].Type,
-				Component: lst[i].Component,
-				Cate:      lst[i].Cate,
-				Name:      dashboard.Name,
-				Tags:      dashboard.Tags,
-				UUID:      dashboard.UUID,
-				Content:   lst[i].Content,
-				CreatedBy: username,
-				UpdatedBy: username,
+				Type:        lst[i].Type,
+				ComponentID: lst[i].ComponentID,
+				Cate:        lst[i].Cate,
+				Name:        dashboard.Name,
+				Tags:        dashboard.Tags,
+				UUID:        dashboard.UUID,
+				Content:     string(contentBytes),
+				CreatedBy:   username,
+				UpdatedBy:   username,
 			}
 
 			if err := bp.Add(rt.Ctx, username); err != nil {
 				reterr[bp.Name] = i18n.Sprintf(c.GetHeader("X-Language"), err.Error())
 			}
 		} else {
+			if lst[i].Type == "collect" {
+				c := make(map[string]interface{})
+				if _, err := toml.Decode(lst[i].Content, &c); err != nil {
+					reterr[lst[i].Name] = err.Error()
+					continue
+				}
+			}
+
 			if err := lst[i].Add(rt.Ctx, username); err != nil {
 				reterr[lst[i].Name] = i18n.Sprintf(c.GetHeader("X-Language"), err.Error())
 			}
@@ -171,19 +192,20 @@ func (rt *Router) builtinPayloadsAdd(c *gin.Context) {
 
 func (rt *Router) builtinPayloadsGets(c *gin.Context) {
 	typ := ginx.QueryStr(c, "type", "")
-	component := ginx.QueryStr(c, "component", "")
+	ComponentID := ginx.QueryInt64(c, "component_id", 0)
+
 	cate := ginx.QueryStr(c, "cate", "")
 	query := ginx.QueryStr(c, "query", "")
 
-	lst, err := models.BuiltinPayloadGets(rt.Ctx, typ, component, cate, query)
+	lst, err := models.BuiltinPayloadGets(rt.Ctx, uint64(ComponentID), typ, cate, query)
 	ginx.NewRender(c).Data(lst, err)
 }
 
 func (rt *Router) builtinPayloadcatesGet(c *gin.Context) {
 	typ := ginx.QueryStr(c, "type", "")
-	component := ginx.QueryStr(c, "component", "")
+	ComponentID := ginx.QueryInt64(c, "component_id", 0)
 
-	cates, err := models.BuiltinPayloadCates(rt.Ctx, typ, component)
+	cates, err := models.BuiltinPayloadCates(rt.Ctx, typ, uint64(ComponentID))
 	ginx.NewRender(c).Data(cates, err)
 }
 
@@ -229,6 +251,11 @@ func (rt *Router) builtinPayloadsPut(c *gin.Context) {
 
 		req.Name = dashboard.Name
 		req.Tags = dashboard.Tags
+	} else if req.Type == "collect" {
+		c := make(map[string]interface{})
+		if _, err := toml.Decode(req.Content, &c); err != nil {
+			ginx.Bomb(http.StatusBadRequest, err.Error())
+		}
 	}
 
 	username := Username(c)
@@ -245,3 +272,15 @@ func (rt *Router) builtinPayloadsDel(c *gin.Context) {
 
 	ginx.NewRender(c).Message(models.BuiltinPayloadDels(rt.Ctx, req.Ids))
 }
+
+func (rt *Router) builtinPayloadsGetByUUIDOrID(c *gin.Context) {
+	uuid := ginx.QueryInt64(c, "uuid", 0)
+	// 优先以 uuid 为准
+	if uuid != 0 {
+		ginx.NewRender(c).Data(models.BuiltinPayloadGet(rt.Ctx, "uuid = ?", uuid))
+		return
+	}
+
+	id := ginx.QueryInt64(c, "id", 0)
+	ginx.NewRender(c).Data(models.BuiltinPayloadGet(rt.Ctx, "id = ?", id))
+}

center/router/router_busi_group.go

@@ -140,3 +140,12 @@ func (rt *Router) busiGroupGet(c *gin.Context) {
 	ginx.Dangerous(bg.FillUserGroups(rt.Ctx))
 	ginx.NewRender(c).Data(bg, nil)
 }
+
+func (rt *Router) busiGroupsGetTags(c *gin.Context) {
+	bgids := str.IdsInt64(ginx.QueryStr(c, "gids", ""), ",")
+	targetIdents, err := models.TargetIndentsGetByBgids(rt.Ctx, bgids)
+	ginx.Dangerous(err)
+	tags, err := models.TargetGetTags(rt.Ctx, targetIdents, true, "busigroup")
+	ginx.Dangerous(err)
+	ginx.NewRender(c).Data(tags, nil)
+}

center/router/router_configs.go

@@ -24,6 +24,11 @@ func (rt *Router) configGet(c *gin.Context) {
 	ginx.NewRender(c).Data(configs, err)
 }
 
+func (rt *Router) configGetAll(c *gin.Context) {
+	config, err := models.ConfigsGetAll(rt.Ctx)
+	ginx.NewRender(c).Data(config, err)
+}
+
 func (rt *Router) configGetByKey(c *gin.Context) {
 	config, err := models.ConfigsGet(rt.Ctx, ginx.QueryStr(c, "key"))
 	ginx.NewRender(c).Data(config, err)

center/router/router_dash_annotation.go

@@ -0,0 +1,99 @@
+package router
+
+import (
+	"fmt"
+	"net/http"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
+
+	"github.com/gin-gonic/gin"
+	"github.com/toolkits/pkg/ginx"
+)
+
+func checkAnnotationPermission(c *gin.Context, ctx *ctx.Context, dashboardId int64) {
+	dashboard, err := models.BoardGetByID(ctx, dashboardId)
+	if err != nil {
+		ginx.Bomb(http.StatusInternalServerError, "failed to get dashboard: %v", err)
+	}
+
+	if dashboard == nil {
+		ginx.Bomb(http.StatusNotFound, "dashboard not found")
+	}
+
+	bg := BusiGroup(ctx, dashboard.GroupId)
+	me := c.MustGet("user").(*models.User)
+	can, err := me.CanDoBusiGroup(ctx, bg, "rw")
+	ginx.Dangerous(err)
+
+	if !can {
+		ginx.Bomb(http.StatusForbidden, "forbidden")
+	}
+}
+
+func (rt *Router) dashAnnotationAdd(c *gin.Context) {
+	var f models.DashAnnotation
+	ginx.BindJSON(c, &f)
+
+	username := c.MustGet("username").(string)
+	now := time.Now().Unix()
+
+	checkAnnotationPermission(c, rt.Ctx, f.DashboardId)
+
+	f.CreateBy = username
+	f.CreateAt = now
+	f.UpdateBy = username
+	f.UpdateAt = now
+
+	ginx.NewRender(c).Data(f.Id, f.Add(rt.Ctx))
+}
+
+func (rt *Router) dashAnnotationGets(c *gin.Context) {
+	dashboardId := ginx.QueryInt64(c, "dashboard_id")
+	from := ginx.QueryInt64(c, "from")
+	to := ginx.QueryInt64(c, "to")
+	limit := ginx.QueryInt(c, "limit", 100)
+
+	lst, err := models.DashAnnotationGets(rt.Ctx, dashboardId, from, to, limit)
+	ginx.NewRender(c).Data(lst, err)
+}
+
+func (rt *Router) dashAnnotationPut(c *gin.Context) {
+	var f models.DashAnnotation
+	ginx.BindJSON(c, &f)
+
+	id := ginx.UrlParamInt64(c, "id")
+	annotation, err := getAnnotationById(rt.Ctx, id)
+	ginx.Dangerous(err)
+
+	checkAnnotationPermission(c, rt.Ctx, annotation.DashboardId)
+
+	f.Id = id
+	f.UpdateAt = time.Now().Unix()
+	f.UpdateBy = c.MustGet("username").(string)
+
+	ginx.NewRender(c).Message(f.Update(rt.Ctx))
+}
+
+func (rt *Router) dashAnnotationDel(c *gin.Context) {
+	id := ginx.UrlParamInt64(c, "id")
+
+	annotation, err := getAnnotationById(rt.Ctx, id)
+	ginx.Dangerous(err)
+	checkAnnotationPermission(c, rt.Ctx, annotation.DashboardId)
+
+	ginx.NewRender(c).Message(models.DashAnnotationDel(rt.Ctx, id))
+}
+
+// 可以提取获取注释的通用方法
+func getAnnotationById(ctx *ctx.Context, id int64) (*models.DashAnnotation, error) {
+	annotation, err := models.DashAnnotationGet(ctx, "id=?", id)
+	if err != nil {
+		return nil, err
+	}
+	if annotation == nil {
+		return nil, fmt.Errorf("annotation not found")
+	}
+	return annotation, nil
+}

center/router/router_datasource.go

@@ -92,10 +92,14 @@ func (rt *Router) datasourceUpsert(c *gin.Context) {
 	var err error
 	var count int64
 
-	err = DatasourceCheck(req)
-	if err != nil {
-		Dangerous(c, err)
-		return
+	if !req.ForceSave {
+		if req.PluginType == models.PROMETHEUS || req.PluginType == models.LOKI || req.PluginType == models.TDENGINE {
+			err = DatasourceCheck(req)
+			if err != nil {
+				Dangerous(c, err)
+				return
+			}
+		}
 	}
 
 	if req.Id == 0 {
@@ -120,12 +124,14 @@ func (rt *Router) datasourceUpsert(c *gin.Context) {
 }
 
 func DatasourceCheck(ds models.Datasource) error {
-	if ds.HTTPJson.Url == "" {
-		return fmt.Errorf("url is empty")
-	}
+	if ds.PluginType == models.PROMETHEUS || ds.PluginType == models.LOKI || ds.PluginType == models.TDENGINE {
+		if ds.HTTPJson.Url == "" {
+			return fmt.Errorf("url is empty")
+		}
 
-	if !strings.HasPrefix(ds.HTTPJson.Url, "http") {
-		return fmt.Errorf("url must start with http or https")
+		if !strings.HasPrefix(ds.HTTPJson.Url, "http") {
+			return fmt.Errorf("url must start with http or https")
+		}
 	}
 
 	client := &http.Client{
@@ -136,11 +142,11 @@ func DatasourceCheck(ds models.Datasource) error {
 		},
 	}
 
-	fullURL := ds.HTTPJson.Url
-	req, err := http.NewRequest("GET", fullURL, nil)
+	var fullURL string
+	req, err := ds.HTTPJson.NewReq(&fullURL)
 	if err != nil {
 		logger.Errorf("Error creating request: %v", err)
-		return fmt.Errorf("request url:%s failed", fullURL)
+		return fmt.Errorf("request urls:%v failed", ds.HTTPJson.GetUrls())
 	}
 
 	if ds.PluginType == models.PROMETHEUS {
@@ -247,3 +253,37 @@ func (rt *Router) getDatasourceIds(c *gin.Context) {
 
 	ginx.NewRender(c).Data(datasourceIds, err)
 }
+
+type datasourceQueryForm struct {
+	Cate              string                   `json:"datasource_cate"`
+	DatasourceQueries []models.DatasourceQuery `json:"datasource_queries"`
+}
+
+type datasourceQueryResp struct {
+	ID   int64  `json:"id"`
+	Name string `json:"name"`
+}
+
+func (rt *Router) datasourceQuery(c *gin.Context) {
+	var dsf datasourceQueryForm
+	ginx.BindJSON(c, &dsf)
+	datasources, err := models.GetDatasourcesGetsByTypes(rt.Ctx, []string{dsf.Cate})
+	ginx.Dangerous(err)
+
+	nameToID := make(map[string]int64)
+	IDToName := make(map[int64]string)
+	for _, ds := range datasources {
+		nameToID[ds.Name] = ds.Id
+		IDToName[ds.Id] = ds.Name
+	}
+
+	ids := models.GetDatasourceIDsByDatasourceQueries(dsf.DatasourceQueries, IDToName, nameToID)
+	var req []datasourceQueryResp
+	for _, id := range ids {
+		req = append(req, datasourceQueryResp{
+			ID:   id,
+			Name: IDToName[id],
+		})
+	}
+	ginx.NewRender(c).Data(req, err)
+}

center/router/router_datasource_db.go

@@ -0,0 +1,101 @@
+package router
+
+import (
+	"context"
+
+	"github.com/ccfos/nightingale/v6/dscache"
+	"github.com/ccfos/nightingale/v6/dskit/types"
+	"github.com/ccfos/nightingale/v6/models"
+
+	"github.com/gin-gonic/gin"
+	"github.com/toolkits/pkg/ginx"
+	"github.com/toolkits/pkg/logger"
+)
+
+func (rt *Router) ShowDatabases(c *gin.Context) {
+	var f models.QueryParam
+	ginx.BindJSON(c, &f)
+
+	plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if !exists {
+		logger.Warningf("cluster:%d not exists", f.DatasourceId)
+		ginx.Bomb(200, "cluster not exists")
+	}
+
+	var databases []string
+	var err error
+	type DatabaseShower interface {
+		ShowDatabases(context.Context) ([]string, error)
+	}
+	switch plug.(type) {
+	case DatabaseShower:
+		databases, err = plug.(DatabaseShower).ShowDatabases(c.Request.Context())
+		ginx.Dangerous(err)
+	default:
+		ginx.Bomb(200, "datasource not exists")
+	}
+
+	if len(databases) == 0 {
+		databases = make([]string, 0)
+	}
+
+	ginx.NewRender(c).Data(databases, nil)
+}
+
+func (rt *Router) ShowTables(c *gin.Context) {
+	var f models.QueryParam
+	ginx.BindJSON(c, &f)
+
+	plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if !exists {
+		logger.Warningf("cluster:%d not exists", f.DatasourceId)
+		ginx.Bomb(200, "cluster not exists")
+	}
+
+	// 只接受一个入参
+	tables := make([]string, 0)
+	var err error
+	type TableShower interface {
+		ShowTables(ctx context.Context, database string) ([]string, error)
+	}
+	switch plug.(type) {
+	case TableShower:
+		if len(f.Querys) > 0 {
+			database, ok := f.Querys[0].(string)
+			if ok {
+				tables, err = plug.(TableShower).ShowTables(c.Request.Context(), database)
+			}
+		}
+	default:
+		ginx.Bomb(200, "datasource not exists")
+	}
+	ginx.NewRender(c).Data(tables, err)
+}
+
+func (rt *Router) DescribeTable(c *gin.Context) {
+	var f models.QueryParam
+	ginx.BindJSON(c, &f)
+
+	plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if !exists {
+		logger.Warningf("cluster:%d not exists", f.DatasourceId)
+		ginx.Bomb(200, "cluster not exists")
+	}
+	// 只接受一个入参
+	columns := make([]*types.ColumnProperty, 0)
+	var err error
+	type TableDescriber interface {
+		DescribeTable(context.Context, interface{}) ([]*types.ColumnProperty, error)
+	}
+	switch plug.(type) {
+	case TableDescriber:
+		client := plug.(TableDescriber)
+		if len(f.Querys) > 0 {
+			columns, err = client.DescribeTable(c.Request.Context(), f.Querys[0])
+		}
+	default:
+		ginx.Bomb(200, "datasource not exists")
+	}
+
+	ginx.NewRender(c).Data(columns, err)
+}

center/router/router_es.go

@@ -0,0 +1,77 @@
+package router
+
+import (
+	"github.com/ccfos/nightingale/v6/datasource/es"
+	"github.com/ccfos/nightingale/v6/dscache"
+
+	"github.com/gin-gonic/gin"
+	"github.com/toolkits/pkg/ginx"
+	"github.com/toolkits/pkg/logger"
+)
+
+type IndexReq struct {
+	Cate         string `json:"cate"`
+	DatasourceId int64  `json:"datasource_id"`
+	Index        string `json:"index"`
+}
+
+type FieldValueReq struct {
+	Cate         string   `json:"cate"`
+	DatasourceId int64    `json:"datasource_id"`
+	Index        string   `json:"index"`
+	Query        FieldObj `json:"query"`
+}
+
+type FieldObj struct {
+	Find  string `json:"find"`
+	Field string `json:"field"`
+	Query string `json:"query"`
+}
+
+func (rt *Router) QueryIndices(c *gin.Context) {
+	var f IndexReq
+	ginx.BindJSON(c, &f)
+
+	plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if !exists {
+		logger.Warningf("cluster:%d not exists", f.DatasourceId)
+		ginx.Bomb(200, "cluster not exists")
+	}
+
+	indices, err := plug.(*es.Elasticsearch).QueryIndices()
+	ginx.Dangerous(err)
+
+	ginx.NewRender(c).Data(indices, nil)
+}
+
+func (rt *Router) QueryFields(c *gin.Context) {
+	var f IndexReq
+	ginx.BindJSON(c, &f)
+
+	plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if !exists {
+		logger.Warningf("cluster:%d not exists", f.DatasourceId)
+		ginx.Bomb(200, "cluster not exists")
+	}
+
+	fields, err := plug.(*es.Elasticsearch).QueryFields([]string{f.Index})
+	ginx.Dangerous(err)
+
+	ginx.NewRender(c).Data(fields, nil)
+}
+
+func (rt *Router) QueryESVariable(c *gin.Context) {
+	var f FieldValueReq
+	ginx.BindJSON(c, &f)
+
+	plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if !exists {
+		logger.Warningf("cluster:%d not exists", f.DatasourceId)
+		ginx.Bomb(200, "cluster not exists")
+	}
+
+	fields, err := plug.(*es.Elasticsearch).QueryFieldValue([]string{f.Index}, f.Query.Field, f.Query.Query)
+	ginx.Dangerous(err)
+
+	ginx.NewRender(c).Data(fields, nil)
+}

center/router/router_funcs.go

@@ -45,6 +45,10 @@ func (rt *Router) statistic(c *gin.Context) {
 		statistics, err = models.ConfigsUserVariableStatistics(rt.Ctx)
 		ginx.NewRender(c).Data(statistics, err)
 		return
+	case "cval":
+		statistics, err = models.ConfigCvalStatistics(rt.Ctx)
+		ginx.NewRender(c).Data(statistics, err)
+		return
 	default:
 		ginx.Bomb(http.StatusBadRequest, "invalid name")
 	}
@@ -65,6 +69,23 @@ func queryDatasourceIds(c *gin.Context) []int64 {
 	return ids
 }
 
+func queryStrListField(c *gin.Context, fieldName string, sep ...string) []string {
+	str := ginx.QueryStr(c, fieldName, "")
+	if str == "" {
+		return nil
+	}
+
+	lst := []string{str}
+	for _, s := range sep {
+		var newLst []string
+		for _, str := range lst {
+			newLst = append(newLst, strings.Split(str, s)...)
+		}
+		lst = newLst
+	}
+	return lst
+}
+
 type idsForm struct {
 	Ids               []int64 `json:"ids"`
 	IsSyncToFlashDuty bool    `json:"is_sync_to_flashduty"`

center/router/router_heartbeat.go

@@ -3,8 +3,10 @@ package router
 import (
 	"compress/gzip"
 	"encoding/json"
-	"fmt"
+	"errors"
 	"io/ioutil"
+	"sort"
+	"strconv"
 	"strings"
 	"time"
 
@@ -57,7 +59,7 @@ func HandleHeartbeat(c *gin.Context, ctx *ctx.Context, engineName string, metaSe
 	}
 
 	if req.Hostname == "" {
-		return req, fmt.Errorf("hostname is required", 400)
+		return req, errors.New("hostname is required")
 	}
 
 	// maybe from pushgw
@@ -79,55 +81,122 @@ func HandleHeartbeat(c *gin.Context, ctx *ctx.Context, engineName string, metaSe
 	identSet.MSet(items)
 
 	if target, has := targetCache.Get(req.Hostname); has && target != nil {
-		gid := ginx.QueryInt64(c, "gid", 0)
+		gidsStr := ginx.QueryStr(c, "gid", "")
+		overwriteGids := ginx.QueryBool(c, "overwrite_gids", false)
 		hostIp := strings.TrimSpace(req.HostIp)
+		gids := strings.Split(gidsStr, ",")
 
-		field := make(map[string]interface{})
-		if gid != 0 && gid != target.GroupId {
-			field["group_id"] = gid
+		if overwriteGids {
+			groupIds := make([]int64, 0)
+			for i := range gids {
+				if gids[i] == "" {
+					continue
+				}
+				groupId, err := strconv.ParseInt(gids[i], 10, 64)
+				if err != nil {
+					logger.Warningf("update target:%s group ids failed, err: %v", req.Hostname, err)
+					continue
+				}
+				groupIds = append(groupIds, groupId)
+			}
+
+			err := models.TargetOverrideBgids(ctx, []string{target.Ident}, groupIds, nil)
+			if err != nil {
+				logger.Warningf("update target:%s group ids failed, err: %v", target.Ident, err)
+			}
+		} else if gidsStr != "" {
+			for i := range gids {
+				groupId, err := strconv.ParseInt(gids[i], 10, 64)
+				if err != nil {
+					logger.Warningf("update target:%s group ids failed, err: %v", req.Hostname, err)
+					continue
+				}
+
+				if !target.MatchGroupId(groupId) {
+					err := models.TargetBindBgids(ctx, []string{target.Ident}, []int64{groupId}, nil)
+					if err != nil {
+						logger.Warningf("update target:%s group ids failed, err: %v", target.Ident, err)
+					}
+				}
+			}
 		}
 
+		newTarget := models.Target{}
+		targetNeedUpdate := false
 		if hostIp != "" && hostIp != target.HostIp {
-			field["host_ip"] = hostIp
+			newTarget.HostIp = hostIp
+			targetNeedUpdate = true
 		}
 
-		tagsMap := target.GetTagsMap()
-		tagNeedUpdate := false
-		for k, v := range req.GlobalLabels {
+		hostTagsMap := target.GetHostTagsMap()
+		hostTagNeedUpdate := false
+		if len(hostTagsMap) != len(req.GlobalLabels) {
+			hostTagNeedUpdate = true
+		} else {
+			for k, v := range req.GlobalLabels {
+				if v == "" {
+					continue
+				}
+
+				if tagv, ok := hostTagsMap[k]; !ok || tagv != v {
+					hostTagNeedUpdate = true
+					break
+				}
+			}
+		}
+
+		if hostTagNeedUpdate {
+			lst := []string{}
+			for k, v := range req.GlobalLabels {
+				lst = append(lst, k+"="+v)
+			}
+			sort.Strings(lst)
+			newTarget.HostTags = lst
+			targetNeedUpdate = true
+		}
+
+		userTagsMap := target.GetTagsMap()
+		userTagNeedUpdate := false
+		userTags := []string{}
+		for k, v := range userTagsMap {
 			if v == "" {
 				continue
 			}
 
-			if tagv, ok := tagsMap[k]; !ok || tagv != v {
-				tagNeedUpdate = true
-				tagsMap[k] = v
+			if _, ok := req.GlobalLabels[k]; !ok {
+				userTags = append(userTags, k+"="+v)
+			} else { // 该key在hostTags中已经存在
+				userTagNeedUpdate = true
 			}
 		}
 
-		if tagNeedUpdate {
-			lst := []string{}
-			for k, v := range tagsMap {
-				lst = append(lst, k+"="+v)
-			}
-			labels := strings.Join(lst, " ") + " "
-			field["tags"] = labels
+		if userTagNeedUpdate {
+			newTarget.Tags = strings.Join(userTags, " ") + " "
+			targetNeedUpdate = true
 		}
 
 		if req.EngineName != "" && req.EngineName != target.EngineName {
-			field["engine_name"] = req.EngineName
+			newTarget.EngineName = req.EngineName
+			targetNeedUpdate = true
 		}
 
 		if req.AgentVersion != "" && req.AgentVersion != target.AgentVersion {
-			field["agent_version"] = req.AgentVersion
+			newTarget.AgentVersion = req.AgentVersion
+			targetNeedUpdate = true
 		}
 
-		if len(field) > 0 {
-			err := target.UpdateFieldsMap(ctx, field)
+		if req.OS != "" && req.OS != target.OS {
+			newTarget.OS = req.OS
+			targetNeedUpdate = true
+		}
+
+		if targetNeedUpdate {
+			err := models.DB(ctx).Model(&target).Updates(newTarget).Error
 			if err != nil {
 				logger.Errorf("update target fields failed, err: %v", err)
 			}
 		}
-		logger.Debugf("heartbeat field:%+v target: %v", field, *target)
+		logger.Debugf("heartbeat field:%+v target: %v", newTarget, *target)
 	}
 
 	return req, nil

center/router/router_mute.go

@@ -50,7 +50,8 @@ func (rt *Router) alertMuteGets(c *gin.Context) {
 	prods := strings.Fields(ginx.QueryStr(c, "prods", ""))
 	bgid := ginx.QueryInt64(c, "bgid", -1)
 	query := ginx.QueryStr(c, "query", "")
-	lst, err := models.AlertMuteGets(rt.Ctx, prods, bgid, query)
+	disabled := ginx.QueryInt(c, "disabled", -1)
+	lst, err := models.AlertMuteGets(rt.Ctx, prods, bgid, disabled, query)
 
 	ginx.NewRender(c).Data(lst, err)
 }

center/router/router_mw.go

@@ -130,9 +130,9 @@ func (rt *Router) User() gin.HandlerFunc {
 
 func (rt *Router) user() gin.HandlerFunc {
 	return func(c *gin.Context) {
-		userid := c.MustGet("userid").(int64)
+		username := c.MustGet("username").(string)
 
-		user, err := models.UserGetById(rt.Ctx, userid)
+		user, err := models.UserGetByUsername(rt.Ctx, username)
 		if err != nil {
 			ginx.Bomb(http.StatusUnauthorized, "unauthorized")
 		}

center/router/router_notification_record.go

@@ -0,0 +1,212 @@
+package router
+
+import (
+	"strings"
+
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
+
+	"github.com/gin-gonic/gin"
+	"github.com/toolkits/pkg/ginx"
+	"github.com/toolkits/pkg/logger"
+)
+
+type NotificationResponse struct {
+	SubRules []SubRule           `json:"sub_rules"`
+	Notifies map[string][]Record `json:"notifies"`
+}
+
+type SubRule struct {
+	SubID    int64               `json:"sub_id"`
+	Notifies map[string][]Record `json:"notifies"`
+}
+
+type Notify struct {
+	Channel string   `json:"channel"`
+	Records []Record `json:"records"`
+}
+
+type Record struct {
+	Target   string `json:"target"`
+	Username string `json:"username"`
+	Status   int    `json:"status"`
+	Detail   string `json:"detail"`
+}
+
+// notificationRecordAdd
+func (rt *Router) notificationRecordAdd(c *gin.Context) {
+	var req []*models.NotificaitonRecord
+	ginx.BindJSON(c, &req)
+	err := models.DB(rt.Ctx).CreateInBatches(req, 100).Error
+	var ids []int64
+	if err == nil {
+		ids = make([]int64, len(req))
+		for i, noti := range req {
+			ids[i] = noti.Id
+		}
+	}
+
+	ginx.NewRender(c).Data(ids, err)
+}
+
+func (rt *Router) notificationRecordList(c *gin.Context) {
+	eid := ginx.UrlParamInt64(c, "eid")
+	lst, err := models.NotificaitonRecordsGetByEventId(rt.Ctx, eid)
+	ginx.Dangerous(err)
+
+	response := buildNotificationResponse(rt.Ctx, lst)
+	ginx.NewRender(c).Data(response, nil)
+}
+
+func buildNotificationResponse(ctx *ctx.Context, nl []*models.NotificaitonRecord) NotificationResponse {
+	response := NotificationResponse{
+		SubRules: []SubRule{},
+		Notifies: make(map[string][]Record),
+	}
+
+	subRuleMap := make(map[int64]*SubRule)
+
+	// Collect all group IDs
+	groupIdSet := make(map[int64]struct{})
+
+	// map[SubId]map[Channel]map[Target]index
+	filter := make(map[int64]map[string]map[string]int)
+
+	for i, n := range nl {
+		// 对相同的 channel-target 进行合并
+		for _, gid := range n.GetGroupIds(ctx) {
+			groupIdSet[gid] = struct{}{}
+		}
+
+		if _, exists := filter[n.SubId]; !exists {
+			filter[n.SubId] = make(map[string]map[string]int)
+		}
+
+		if _, exists := filter[n.SubId][n.Channel]; !exists {
+			filter[n.SubId][n.Channel] = make(map[string]int)
+		}
+
+		idx, exists := filter[n.SubId][n.Channel][n.Target]
+		if !exists {
+			filter[n.SubId][n.Channel][n.Target] = i
+		} else {
+			if nl[idx].Status < n.Status {
+				nl[idx].Status = n.Status
+			}
+			nl[idx].Details = nl[idx].Details + ", " + n.Details
+			nl[i] = nil
+		}
+
+	}
+
+	// Fill usernames only once
+	usernameByTarget := fillUserNames(ctx, groupIdSet)
+
+	for _, n := range nl {
+		if n == nil {
+			continue
+		}
+
+		m := usernameByTarget[n.Target]
+		usernames := make([]string, 0, len(m))
+		for k := range m {
+			usernames = append(usernames, k)
+		}
+
+		if !checkChannel(n.Channel) {
+			// Hide sensitive information
+			n.Target = replaceLastEightChars(n.Target)
+		}
+		record := Record{
+			Target: n.Target,
+			Status: n.Status,
+			Detail: n.Details,
+		}
+
+		record.Username = strings.Join(usernames, ",")
+
+		if n.SubId > 0 {
+			// Handle SubRules
+			subRule, ok := subRuleMap[n.SubId]
+			if !ok {
+				newSubRule := &SubRule{
+					SubID: n.SubId,
+				}
+				newSubRule.Notifies = make(map[string][]Record)
+				newSubRule.Notifies[n.Channel] = []Record{record}
+
+				subRuleMap[n.SubId] = newSubRule
+			} else {
+				if _, exists := subRule.Notifies[n.Channel]; !exists {
+
+					subRule.Notifies[n.Channel] = []Record{record}
+				} else {
+					subRule.Notifies[n.Channel] = append(subRule.Notifies[n.Channel], record)
+				}
+			}
+			continue
+		}
+
+		if response.Notifies == nil {
+			response.Notifies = make(map[string][]Record)
+		}
+
+		if _, exists := response.Notifies[n.Channel]; !exists {
+			response.Notifies[n.Channel] = []Record{record}
+		} else {
+			response.Notifies[n.Channel] = append(response.Notifies[n.Channel], record)
+		}
+	}
+
+	for _, subRule := range subRuleMap {
+		response.SubRules = append(response.SubRules, *subRule)
+	}
+
+	return response
+}
+
+// check channel is one of the following:  tx-sms, tx-voice, ali-sms, ali-voice, email, script
+func checkChannel(channel string) bool {
+	switch channel {
+	case "tx-sms", "tx-voice", "ali-sms", "ali-voice", "email", "script":
+		return true
+	}
+	return false
+}
+
+func replaceLastEightChars(s string) string {
+	if len(s) <= 8 {
+		return strings.Repeat("*", len(s))
+	}
+	return s[:len(s)-8] + strings.Repeat("*", 8)
+}
+
+func fillUserNames(ctx *ctx.Context, groupIdSet map[int64]struct{}) map[string]map[string]struct{} {
+	userNameByTarget := make(map[string]map[string]struct{})
+
+	gids := make([]int64, 0, len(groupIdSet))
+	for gid := range groupIdSet {
+		gids = append(gids, gid)
+	}
+
+	users, err := models.UsersGetByGroupIds(ctx, gids)
+	if err != nil {
+		logger.Errorf("UsersGetByGroupIds failed, err: %v", err)
+		return userNameByTarget
+	}
+
+	for _, user := range users {
+		logger.Warningf("user: %s", user.Username)
+		for _, ch := range models.DefaultChannels {
+			target, exist := user.ExtractToken(ch)
+			if exist {
+				if _, ok := userNameByTarget[target]; !ok {
+					userNameByTarget[target] = make(map[string]struct{})
+				}
+				userNameByTarget[target][user.Username] = struct{}{}
+			}
+		}
+	}
+
+	return userNameByTarget
+}

center/router/router_notify_config.go

@@ -182,7 +182,7 @@ func (rt *Router) notifyConfigPut(c *gin.Context) {
 
 		smtp, errSmtp := SmtpValidate(text)
 		ginx.Dangerous(errSmtp)
-		go sender.RestartEmailSender(smtp)
+		go sender.RestartEmailSender(rt.Ctx, smtp)
 	}
 
 	ginx.NewRender(c).Message(nil)

center/router/router_notify_tpl.go

@@ -68,7 +68,7 @@ func (rt *Router) notifyTplUpdate(c *gin.Context) {
 	}
 
 	// get the count of the same channel and name but different id
-	count, err := models.Count(models.DB(rt.Ctx).Model(&models.NotifyTpl{}).Where("channel = ? or name = ? and id <> ?", f.Channel, f.Name, f.Id))
+	count, err := models.Count(models.DB(rt.Ctx).Model(&models.NotifyTpl{}).Where("(channel = ? or name = ?) and id <> ?", f.Channel, f.Name, f.Id))
 	ginx.Dangerous(err)
 	if count != 0 {
 		ginx.Bomb(200, "Refuse to create duplicate channel or name")
@@ -138,7 +138,7 @@ func (rt *Router) notifyTplPreview(c *gin.Context) {
 			continue
 		}
 
-		arr := strings.Split(pair, "=")
+		arr := strings.SplitN(pair, "=", 2)
 		if len(arr) != 2 {
 			continue
 		}
@@ -161,7 +161,11 @@ func (rt *Router) notifyTplPreview(c *gin.Context) {
 func (rt *Router) notifyTplAdd(c *gin.Context) {
 	var f models.NotifyTpl
 	ginx.BindJSON(c, &f)
-	f.Channel = strings.TrimSpace(f.Channel)
+
+        user := c.MustGet("user").(*models.User)
+        f.CreateBy = user.Username
+	
+        f.Channel = strings.TrimSpace(f.Channel)
 	ginx.Dangerous(templateValidate(f))
 
 	count, err := models.Count(models.DB(rt.Ctx).Model(&models.NotifyTpl{}).Where("channel = ? or name = ?", f.Channel, f.Name))
@@ -169,6 +173,8 @@ func (rt *Router) notifyTplAdd(c *gin.Context) {
 	if count != 0 {
 		ginx.Bomb(200, "Refuse to create duplicate channel(unique)")
 	}
+
+        f.CreateAt = time.Now().Unix()
 	ginx.NewRender(c).Message(f.Create(rt.Ctx))
 }
 

center/router/router_proxy.go

@@ -7,7 +7,6 @@ import (
 	"net"
 	"net/http"
 	"net/http/httputil"
-	"net/url"
 	"strings"
 	"sync"
 	"time"
@@ -112,9 +111,9 @@ func (rt *Router) dsProxy(c *gin.Context) {
 		return
 	}
 
-	target, err := url.Parse(ds.HTTPJson.Url)
+	target, err := ds.HTTPJson.ParseUrl()
 	if err != nil {
-		c.String(http.StatusInternalServerError, "invalid  url: %s", ds.HTTPJson.Url)
+		c.String(http.StatusInternalServerError, "invalid urls: %s", ds.HTTPJson.GetUrls())
 		return
 	}
 

center/router/router_query.go

@@ -0,0 +1,181 @@
+package router
+
+import (
+	"fmt"
+	"sort"
+
+	"github.com/ccfos/nightingale/v6/dscache"
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/gin-gonic/gin"
+	"github.com/toolkits/pkg/ginx"
+	"github.com/toolkits/pkg/logger"
+)
+
+func CheckDsPerm(c *gin.Context, dsId int64, cate string, q interface{}) bool {
+	// todo: 后续需要根据 cate 判断是否需要权限
+	return true
+}
+
+type QueryFrom struct {
+	Queries []Query `json:"queries"`
+	Exps    []Exp   `json:"exps"`
+}
+
+type Query struct {
+	Ref    string      `json:"ref"`
+	Did    int64       `json:"ds_id"`
+	DsCate string      `json:"ds_cate"`
+	Query  interface{} `json:"query"`
+}
+
+type Exp struct {
+	Exp string `json:"exp"`
+	Ref string `json:"ref"`
+}
+
+type LogResp struct {
+	Total int64         `json:"total"`
+	List  []interface{} `json:"list"`
+}
+
+func (rt *Router) QueryLogBatch(c *gin.Context) {
+	var f QueryFrom
+	ginx.BindJSON(c, &f)
+
+	var resp LogResp
+	var errMsg string
+	for _, q := range f.Queries {
+		if !rt.Center.AnonymousAccess.PromQuerier && !CheckDsPerm(c, q.Did, q.DsCate, q) {
+			ginx.Bomb(200, "no permission")
+		}
+
+		plug, exists := dscache.DsCache.Get(q.DsCate, q.Did)
+		if !exists {
+			logger.Warningf("cluster:%d not exists query:%+v", q.Did, q)
+			ginx.Bomb(200, "cluster not exists")
+		}
+
+		data, total, err := plug.QueryLog(c.Request.Context(), q.Query)
+		if err != nil {
+			errMsg += fmt.Sprintf("query data error: %v query:%v\n ", err, q)
+			logger.Warningf("query data error: %v query:%v", err, q)
+			continue
+		}
+
+		m := make(map[string]interface{})
+		m["ref"] = q.Ref
+		m["ds_id"] = q.Did
+		m["ds_cate"] = q.DsCate
+		m["data"] = data
+		resp.List = append(resp.List, m)
+		resp.Total += total
+	}
+
+	if errMsg != "" || len(resp.List) == 0 {
+		ginx.Bomb(200, errMsg)
+	}
+
+	ginx.NewRender(c).Data(resp, nil)
+}
+
+func (rt *Router) QueryData(c *gin.Context) {
+	var f models.QueryParam
+	ginx.BindJSON(c, &f)
+
+	var resp []models.DataResp
+	var err error
+	for _, q := range f.Querys {
+		if !rt.Center.AnonymousAccess.PromQuerier && !CheckDsPerm(c, f.DatasourceId, f.Cate, q) {
+			ginx.Bomb(403, "no permission")
+		}
+
+		plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+		if !exists {
+			logger.Warningf("cluster:%d not exists", f.DatasourceId)
+			ginx.Bomb(200, "cluster not exists")
+		}
+		var datas []models.DataResp
+		datas, err = plug.QueryData(c.Request.Context(), q)
+		if err != nil {
+			logger.Warningf("query data error: req:%+v err:%v", q, err)
+			ginx.Bomb(200, "err:%v", err)
+		}
+		logger.Debugf("query data: req:%+v resp:%+v", q, datas)
+		resp = append(resp, datas...)
+	}
+	// 面向API的统一处理
+	// 按照 .Metric 排序
+	// 确保仪表盘中相同图例的曲线颜色相同
+	if len(resp) > 1 {
+		sort.Slice(resp, func(i, j int) bool {
+			if resp[i].Metric != nil && resp[j].Metric != nil {
+				return resp[i].Metric.String() < resp[j].Metric.String()
+			}
+			return false
+		})
+	}
+
+	ginx.NewRender(c).Data(resp, err)
+}
+
+func (rt *Router) QueryLogV2(c *gin.Context) {
+	var f models.QueryParam
+	ginx.BindJSON(c, &f)
+
+	var resp LogResp
+	var errMsg string
+	for _, q := range f.Querys {
+		if !rt.Center.AnonymousAccess.PromQuerier && !CheckDsPerm(c, f.DatasourceId, f.Cate, q) {
+			ginx.Bomb(200, "no permission")
+		}
+
+		plug, exists := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+		if !exists {
+			logger.Warningf("cluster:%d not exists query:%+v", f.DatasourceId, f)
+			ginx.Bomb(200, "cluster not exists")
+		}
+
+		data, total, err := plug.QueryLog(c.Request.Context(), q)
+		if err != nil {
+			errMsg += fmt.Sprintf("query data error: %v query:%v\n ", err, q)
+			logger.Warningf("query data error: %v query:%v", err, q)
+			continue
+		}
+		resp.List = append(resp.List, data...)
+		resp.Total += total
+	}
+
+	if errMsg != "" || len(resp.List) == 0 {
+		ginx.Bomb(200, errMsg)
+	}
+
+	ginx.NewRender(c).Data(resp, nil)
+}
+
+func (rt *Router) QueryLog(c *gin.Context) {
+	var f models.QueryParam
+	ginx.BindJSON(c, &f)
+
+	var resp []interface{}
+	for _, q := range f.Querys {
+		if !rt.Center.AnonymousAccess.PromQuerier && !CheckDsPerm(c, f.DatasourceId, f.Cate, q) {
+			ginx.Bomb(200, "no permission")
+		}
+
+		plug, exists := dscache.DsCache.Get("elasticsearch", f.DatasourceId)
+		if !exists {
+			logger.Warningf("cluster:%d not exists", f.DatasourceId)
+			ginx.Bomb(200, "cluster not exists")
+		}
+
+		data, _, err := plug.QueryLog(c.Request.Context(), q)
+		if err != nil {
+			logger.Warningf("query data error: %v", err)
+			ginx.Bomb(200, "err:%v", err)
+			continue
+		}
+		resp = append(resp, data...)
+	}
+
+	ginx.NewRender(c).Data(resp, nil)
+}

center/router/router_recording_rule.go

@@ -3,8 +3,6 @@ package router
 import (
 	"encoding/json"
 	"net/http"
-	"strconv"
-	"strings"
 	"time"
 
 	"github.com/ccfos/nightingale/v6/models"
@@ -74,6 +72,14 @@ func (rt *Router) recordingRuleAddByFE(c *gin.Context) {
 		ginx.Bomb(http.StatusBadRequest, "input json is empty")
 	}
 
+	for i := range lst {
+		if len(lst[i].DatasourceQueries) == 0 {
+			lst[i].DatasourceQueries = []models.DatasourceQuery{
+				models.DataSourceQueryAll,
+			}
+		}
+	}
+
 	bgid := ginx.UrlParamInt64(c, "id")
 	reterr := make(map[string]string)
 	for i := 0; i < count; i++ {
@@ -137,23 +143,10 @@ func (rt *Router) recordingRulePutFields(c *gin.Context) {
 	f.Fields["update_by"] = c.MustGet("username").(string)
 	f.Fields["update_at"] = time.Now().Unix()
 
-	if _, ok := f.Fields["datasource_ids"]; ok {
-		// datasource_ids = "1 2 3"
-		idsStr := strings.Fields(f.Fields["datasource_ids"].(string))
-		ids := make([]int64, 0)
-		for _, idStr := range idsStr {
-			id, err := strconv.ParseInt(idStr, 10, 64)
-			if err != nil {
-				ginx.Bomb(http.StatusBadRequest, "datasource_ids error")
-			}
-			ids = append(ids, id)
-		}
-
-		bs, err := json.Marshal(ids)
-		if err != nil {
-			ginx.Bomb(http.StatusBadRequest, "datasource_ids error")
-		}
-		f.Fields["datasource_ids"] = string(bs)
+	if datasourceQueries, ok := f.Fields["datasource_queries"]; ok {
+		bytes, err := json.Marshal(datasourceQueries)
+		ginx.Dangerous(err)
+		f.Fields["datasource_queries"] = string(bytes)
 	}
 
 	for i := 0; i < len(f.Ids); i++ {

center/router/router_target.go

@@ -9,6 +9,7 @@ import (
 	"time"
 
 	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/storage"
 
 	"github.com/gin-gonic/gin"
@@ -52,6 +53,8 @@ func (rt *Router) targetGets(c *gin.Context) {
 	order := ginx.QueryStr(c, "order", "ident")
 	desc := ginx.QueryBool(c, "desc", false)
 
+	hosts := queryStrListField(c, "hosts", ",", " ", "\n")
+
 	var err error
 	if len(bgids) == 0 {
 		user := c.MustGet("user").(*models.User)
@@ -65,11 +68,13 @@ func (rt *Router) targetGets(c *gin.Context) {
 			bgids = append(bgids, 0)
 		}
 	}
+
 	options := []models.BuildTargetWhereOption{
 		models.BuildTargetWhereWithBgids(bgids),
 		models.BuildTargetWhereWithDsIds(dsIds),
 		models.BuildTargetWhereWithQuery(query),
 		models.BuildTargetWhereWithDowntime(downtime),
+		models.BuildTargetWhereWithHosts(hosts),
 	}
 	total, err := models.TargetTotal(rt.Ctx, options...)
 	ginx.Dangerous(err)
@@ -78,6 +83,13 @@ func (rt *Router) targetGets(c *gin.Context) {
 		ginx.Offset(c, limit), order, desc, options...)
 	ginx.Dangerous(err)
 
+	tgs, err := models.TargetBusiGroupsGetAll(rt.Ctx)
+	ginx.Dangerous(err)
+
+	for _, t := range list {
+		t.GroupIds = tgs[t.Ident]
+	}
+
 	if err == nil {
 		now := time.Now()
 		cache := make(map[int64]*models.BusiGroup)
@@ -157,7 +169,8 @@ func (rt *Router) targetGetsByService(c *gin.Context) {
 func (rt *Router) targetGetTags(c *gin.Context) {
 	idents := ginx.QueryStr(c, "idents", "")
 	idents = strings.ReplaceAll(idents, ",", " ")
-	lst, err := models.TargetGetTags(rt.Ctx, strings.Fields(idents))
+	ignoreHostTag := ginx.QueryBool(c, "ignore_host_tag", false)
+	lst, err := models.TargetGetTags(rt.Ctx, strings.Fields(idents), ignoreHostTag, "")
 	ginx.NewRender(c).Data(lst, err)
 }
 
@@ -262,7 +275,7 @@ func (rt *Router) validateTags(tags []string) error {
 func (rt *Router) addTagsToTarget(target *models.Target, tags []string) error {
 	for _, tag := range tags {
 		tagKey := strings.Split(tag, "=")[0]
-		if strings.Contains(target.Tags, tagKey+"=") {
+		if _, exist := target.TagsMap[tagKey]; exist {
 			return fmt.Errorf("duplicate tagkey(%s)", tagKey)
 		}
 	}
@@ -379,8 +392,31 @@ type targetBgidForm struct {
 	Bgid    int64    `json:"bgid"`
 }
 
-func (rt *Router) targetUpdateBgid(c *gin.Context) {
-	var f targetBgidForm
+type targetBgidsForm struct {
+	Idents  []string `json:"idents" binding:"required_without=HostIps"`
+	HostIps []string `json:"host_ips" binding:"required_without=Idents"`
+	Bgids   []int64  `json:"bgids"`
+	Tags    []string `json:"tags"`
+	Action  string   `json:"action"` // add del reset
+}
+
+func haveNeverGroupedIdent(ctx *ctx.Context, idents []string) (bool, error) {
+	for _, ident := range idents {
+		bgids, err := models.TargetGroupIdsGetByIdent(ctx, ident)
+		if err != nil {
+			return false, err
+		}
+
+		if len(bgids) <= 0 {
+			return true, nil
+		}
+	}
+
+	return false, nil
+}
+
+func (rt *Router) targetBindBgids(c *gin.Context) {
+	var f targetBgidsForm
 	var err error
 	var failedResults = make(map[string]string)
 	ginx.BindJSON(c, &f)
@@ -396,35 +432,24 @@ func (rt *Router) targetUpdateBgid(c *gin.Context) {
 	}
 
 	user := c.MustGet("user").(*models.User)
-	if user.IsAdmin() {
-		ginx.NewRender(c).Data(failedResults, models.TargetUpdateBgid(rt.Ctx, f.Idents, f.Bgid, false))
-		return
-	}
-
-	if f.Bgid > 0 {
-		// 把要操作的机器分成两部分，一部分是bgid为0，需要管理员分配，另一部分bgid>0，说明是业务组内部想调整
-		// 比如原来分配给didiyun的机器，didiyun的管理员想把部分机器调整到didiyun-ceph下
-		// 对于调整的这种情况，当前登录用户要对这批机器有操作权限，同时还要对目标BG有操作权限
-		orphans, err := models.IdentsFilter(rt.Ctx, f.Idents, "group_id = ?", 0)
+	if !user.IsAdmin() {
+		// 普通用户，检查用户是否有权限操作所有请求的业务组
+		existing, _, err := models.SeparateTargetIdents(rt.Ctx, f.Idents)
 		ginx.Dangerous(err)
+		rt.checkTargetPerm(c, existing)
 
-		// 机器里边存在未归组的，登录用户就需要是admin
-		if len(orphans) > 0 && !user.IsAdmin() {
-			can, err := user.CheckPerm(rt.Ctx, "/targets/bind")
+		var groupIds []int64
+		if f.Action == "reset" {
+			// 如果是复写，则需要检查用户是否有权限操作机器之前的业务组
+			bgids, err := models.TargetGroupIdsGetByIdents(rt.Ctx, f.Idents)
 			ginx.Dangerous(err)
-			if !can {
-				ginx.Bomb(http.StatusForbidden, "No permission. Only admin can assign BG")
-			}
+
+			groupIds = append(groupIds, bgids...)
 		}
+		groupIds = append(groupIds, f.Bgids...)
 
-		reBelongs, err := models.IdentsFilter(rt.Ctx, f.Idents, "group_id > ?", 0)
-		ginx.Dangerous(err)
-
-		if len(reBelongs) > 0 {
-			// 对于这些要重新分配的机器，操作者要对这些机器本身有权限，同时要对目标bgid有权限
-			rt.checkTargetPerm(c, f.Idents)
-
-			bg := BusiGroup(rt.Ctx, f.Bgid)
+		for _, bgid := range groupIds {
+			bg := BusiGroup(rt.Ctx, bgid)
 			can, err := user.CanDoBusiGroup(rt.Ctx, bg, "rw")
 			ginx.Dangerous(err)
 
@@ -432,14 +457,28 @@ func (rt *Router) targetUpdateBgid(c *gin.Context) {
 				ginx.Bomb(http.StatusForbidden, "No permission. You are not admin of BG(%s)", bg.Name)
 			}
 		}
-	} else if f.Bgid == 0 {
-		// 退还机器
-		rt.checkTargetPerm(c, f.Idents)
-	} else {
-		ginx.Bomb(http.StatusBadRequest, "invalid bgid")
+		isNeverGrouped, checkErr := haveNeverGroupedIdent(rt.Ctx, f.Idents)
+		ginx.Dangerous(checkErr)
+
+		if isNeverGrouped {
+			can, err := user.CheckPerm(rt.Ctx, "/targets/bind")
+			ginx.Dangerous(err)
+			if !can {
+				ginx.Bomb(http.StatusForbidden, "No permission. Only admin can assign BG")
+			}
+		}
 	}
 
-	ginx.NewRender(c).Data(failedResults, models.TargetUpdateBgid(rt.Ctx, f.Idents, f.Bgid, false))
+	switch f.Action {
+	case "add":
+		ginx.NewRender(c).Data(failedResults, models.TargetBindBgids(rt.Ctx, f.Idents, f.Bgids, f.Tags))
+	case "del":
+		ginx.NewRender(c).Data(failedResults, models.TargetUnbindBgids(rt.Ctx, f.Idents, f.Bgids))
+	case "reset":
+		ginx.NewRender(c).Data(failedResults, models.TargetOverrideBgids(rt.Ctx, f.Idents, f.Bgids, f.Tags))
+	default:
+		ginx.Bomb(http.StatusBadRequest, "invalid action")
+	}
 }
 
 func (rt *Router) targetUpdateBgidByService(c *gin.Context) {
@@ -458,7 +497,7 @@ func (rt *Router) targetUpdateBgidByService(c *gin.Context) {
 		ginx.Bomb(http.StatusBadRequest, err.Error())
 	}
 
-	ginx.NewRender(c).Data(failedResults, models.TargetUpdateBgid(rt.Ctx, f.Idents, f.Bgid, false))
+	ginx.NewRender(c).Data(failedResults, models.TargetOverrideBgids(rt.Ctx, f.Idents, []int64{f.Bgid}, nil))
 }
 
 type identsForm struct {
@@ -482,7 +521,7 @@ func (rt *Router) targetDel(c *gin.Context) {
 		ginx.Bomb(http.StatusBadRequest, err.Error())
 	}
 
-	ginx.NewRender(c).Data(failedResults, models.TargetDel(rt.Ctx, f.Idents))
+	ginx.NewRender(c).Data(failedResults, models.TargetDel(rt.Ctx, f.Idents, rt.TargetDeleteHook))
 }
 
 func (rt *Router) targetDelByService(c *gin.Context) {
@@ -501,7 +540,7 @@ func (rt *Router) targetDelByService(c *gin.Context) {
 		ginx.Bomb(http.StatusBadRequest, err.Error())
 	}
 
-	ginx.NewRender(c).Data(failedResults, models.TargetDel(rt.Ctx, f.Idents))
+	ginx.NewRender(c).Data(failedResults, models.TargetDel(rt.Ctx, f.Idents, rt.TargetDeleteHook))
 }
 
 func (rt *Router) checkTargetPerm(c *gin.Context, idents []string) {
@@ -531,3 +570,18 @@ func (rt *Router) targetsOfAlertRule(c *gin.Context) {
 
 	ginx.NewRender(c).Data(ret, err)
 }
+
+func (rt *Router) targetsOfHostQuery(c *gin.Context) {
+	var queries []models.HostQuery
+	ginx.BindJSON(c, &queries)
+
+	hostsQuery := models.GetHostsQuery(queries)
+	session := models.TargetFilterQueryBuild(rt.Ctx, hostsQuery, 0, 0)
+	var lst []*models.Target
+	err := session.Find(&lst).Error
+	if err != nil {
+		ginx.Bomb(http.StatusInternalServerError, err.Error())
+	}
+
+	ginx.NewRender(c).Data(lst, nil)
+}

center/router/router_task_tpl.go

@@ -196,6 +196,13 @@ func (rt *Router) taskTplDel(c *gin.Context) {
 		return
 	}
 
+	ids, err := models.GetAlertRuleIdsByTaskId(rt.Ctx, tid)
+	ginx.Dangerous(err)
+	if len(ids) > 0 {
+		ginx.NewRender(c).Message("can't del this task tpl, used by alert rule ids(%v) ", ids)
+		return
+	}
+
 	ginx.NewRender(c).Message(tpl.Del(rt.Ctx))
 }
 

center/router/router_tdengine.go

@@ -1,14 +1,14 @@
 package router
 
 import (
-	"net/http"
-
+	"fmt"
 	"github.com/ccfos/nightingale/v6/center/cconf"
+	"github.com/ccfos/nightingale/v6/datasource/tdengine"
+	"github.com/ccfos/nightingale/v6/dscache"
 	"github.com/ccfos/nightingale/v6/models"
-
 	"github.com/gin-gonic/gin"
 	"github.com/toolkits/pkg/ginx"
-	"github.com/toolkits/pkg/logger"
+	"net/http"
 )
 
 type databasesQueryForm struct {
@@ -20,13 +20,13 @@ func (rt *Router) tdengineDatabases(c *gin.Context) {
 	var f databasesQueryForm
 	ginx.BindJSON(c, &f)
 
-	tdClient := rt.TdendgineClients.GetCli(f.DatasourceId)
-	if tdClient == nil {
+	datasource, hit := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if _, ok := datasource.(*tdengine.TDengine); !hit || !ok {
 		ginx.NewRender(c, http.StatusNotFound).Message("No such datasource")
 		return
 	}
 
-	databases, err := tdClient.GetDatabases()
+	databases, err := datasource.(*tdengine.TDengine).ShowDatabases(rt.Ctx.Ctx)
 	ginx.NewRender(c).Data(databases, err)
 }
 
@@ -37,18 +37,29 @@ type tablesQueryForm struct {
 	IsStable     bool   `json:"is_stable"`
 }
 
+type Column struct {
+	Name string `json:"name"`
+	Type string `json:"type"`
+	Size int    `json:"size"`
+}
+
 // get tdengine tables
 func (rt *Router) tdengineTables(c *gin.Context) {
 	var f tablesQueryForm
 	ginx.BindJSON(c, &f)
 
-	tdClient := rt.TdendgineClients.GetCli(f.DatasourceId)
-	if tdClient == nil {
+	datasource, hit := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if _, ok := datasource.(*tdengine.TDengine); !hit || !ok {
 		ginx.NewRender(c, http.StatusNotFound).Message("No such datasource")
 		return
 	}
 
-	tables, err := tdClient.GetTables(f.Database, f.IsStable)
+	database := fmt.Sprintf("%s.tables", f.Database)
+	if f.IsStable {
+		database = fmt.Sprintf("%s.stables", f.Database)
+	}
+
+	tables, err := datasource.(*tdengine.TDengine).ShowTables(rt.Ctx.Ctx, database)
 	ginx.NewRender(c).Data(tables, err)
 }
 
@@ -59,50 +70,31 @@ type columnsQueryForm struct {
 	Table        string `json:"table"`
 }
 
-// get tdengine columns
 func (rt *Router) tdengineColumns(c *gin.Context) {
 	var f columnsQueryForm
 	ginx.BindJSON(c, &f)
 
-	tdClient := rt.TdendgineClients.GetCli(f.DatasourceId)
-	if tdClient == nil {
+	datasource, hit := dscache.DsCache.Get(f.Cate, f.DatasourceId)
+	if _, ok := datasource.(*tdengine.TDengine); !hit || !ok {
 		ginx.NewRender(c, http.StatusNotFound).Message("No such datasource")
 		return
 	}
 
-	columns, err := tdClient.GetColumns(f.Database, f.Table)
-	ginx.NewRender(c).Data(columns, err)
-}
-
-func (rt *Router) QueryData(c *gin.Context) {
-	var f models.QueryParam
-	ginx.BindJSON(c, &f)
-
-	var resp []models.DataResp
-	var err error
-	tdClient := rt.TdendgineClients.GetCli(f.DatasourceId)
-	for _, q := range f.Querys {
-		datas, err := tdClient.Query(q)
-		ginx.Dangerous(err)
-		resp = append(resp, datas...)
+	query := map[string]string{
+		"database": f.Database,
+		"table":    f.Table,
 	}
 
-	ginx.NewRender(c).Data(resp, err)
-}
-
-func (rt *Router) QueryLog(c *gin.Context) {
-	var f models.QueryParam
-	ginx.BindJSON(c, &f)
-
-	tdClient := rt.TdendgineClients.GetCli(f.DatasourceId)
-	if len(f.Querys) == 0 {
-		ginx.Bomb(200, "querys is empty")
-		return
+	columns, err := datasource.(*tdengine.TDengine).DescribeTable(rt.Ctx.Ctx, query)
+	// 对齐前端，后续可以将 tdEngine 的查数据的接口都统一
+	tdColumns := make([]Column, len(columns))
+	for i, column := range columns {
+		tdColumns[i] = Column{
+			Name: column.Field,
+			Type: column.Type,
+		}
 	}
-
-	data, err := tdClient.QueryLog(f.Querys[0])
-	logger.Debugf("tdengine query:%s result: %+v", f.Querys[0], data)
-	ginx.NewRender(c).Data(data, err)
+	ginx.NewRender(c).Data(tdColumns, err)
 }
 
 // query sql template

center/router/router_user.go

@@ -48,7 +48,7 @@ func (rt *Router) userGets(c *gin.Context) {
 	order := ginx.QueryStr(c, "order", "username")
 	desc := ginx.QueryBool(c, "desc", false)
 
-	rt.UserCache.UpdateUsersLastActiveTime()
+	go rt.UserCache.UpdateUsersLastActiveTime()
 	total, err := models.UserTotal(rt.Ctx, query, stime, etime)
 	ginx.Dangerous(err)
 

center/sso/init.go

@@ -1,6 +1,7 @@
 package sso
 
 import (
+	"fmt"
 	"log"
 	"time"
 
@@ -145,7 +146,7 @@ func Init(center cconf.Center, ctx *ctx.Context, configCache *memsto.ConfigCache
 		}
 	}
 	if configCache == nil {
-		logger.Error("configCache is nil, sso initialization failed")
+		log.Fatalln(fmt.Errorf("configCache is nil, sso initialization failed"))
 	}
 	ssoClient.configCache = configCache
 	userVariableMap := configCache.Get()

cli/upgrade/config.go

@@ -33,7 +33,7 @@ type ClusterOptions struct {
 	MaxIdleConnsPerHost int
 }
 
-func Parse(fpath string, configPtr interface{}) error {
+func Parse(fpath string, configPtr *Config) error {
 	var (
 		tBuf []byte
 	)

cmd/edge/edge.go

@@ -5,8 +5,11 @@ import (
 	"errors"
 	"fmt"
 
+	"github.com/ccfos/nightingale/v6/dscache"
+
 	"github.com/ccfos/nightingale/v6/alert"
 	"github.com/ccfos/nightingale/v6/alert/astats"
+	"github.com/ccfos/nightingale/v6/alert/dispatch"
 	"github.com/ccfos/nightingale/v6/alert/process"
 	alertrt "github.com/ccfos/nightingale/v6/alert/router"
 	"github.com/ccfos/nightingale/v6/center/metas"
@@ -16,13 +19,12 @@ import (
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/pkg/httpx"
 	"github.com/ccfos/nightingale/v6/pkg/logx"
+	"github.com/ccfos/nightingale/v6/pkg/macros"
 	"github.com/ccfos/nightingale/v6/prom"
 	"github.com/ccfos/nightingale/v6/pushgw/idents"
 	pushgwrt "github.com/ccfos/nightingale/v6/pushgw/router"
 	"github.com/ccfos/nightingale/v6/pushgw/writer"
 	"github.com/ccfos/nightingale/v6/storage"
-	"github.com/ccfos/nightingale/v6/tdengine"
-
 	"github.com/flashcatcloud/ibex/src/cmd/ibex"
 )
 
@@ -52,12 +54,16 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 
 	targetCache := memsto.NewTargetCache(ctx, syncStats, redis)
 	busiGroupCache := memsto.NewBusiGroupCache(ctx, syncStats)
+	configCvalCache := memsto.NewCvalCache(ctx, syncStats)
 	idents := idents.New(ctx, redis)
 	metas := metas.New(redis)
 	writers := writer.NewWriters(config.Pushgw)
 	pushgwRouter := pushgwrt.New(config.HTTP, config.Pushgw, config.Alert, targetCache, busiGroupCache, idents, metas, writers, ctx)
-	r := httpx.GinEngine(config.Global.RunMode, config.HTTP)
+	r := httpx.GinEngine(config.Global.RunMode, config.HTTP, configCvalCache.PrintBodyPaths, configCvalCache.PrintAccessLog)
+
 	pushgwRouter.Config(r)
+	macros.RegisterMacro(macros.MacroInVain)
+	dscache.Init(ctx, false)
 
 	if !config.Alert.Disable {
 		configCache := memsto.NewConfigCache(ctx, syncStats, nil, "")
@@ -71,11 +77,13 @@ func Initialize(configDir string, cryptoKey string) (func(), error) {
 		taskTplsCache := memsto.NewTaskTplCache(ctx)
 
 		promClients := prom.NewPromClient(ctx)
-		tdengineClients := tdengine.NewTdengineClient(ctx, config.Alert.Heartbeat)
+
+		dispatch.InitRegisterQueryFunc(promClients)
+
 		externalProcessors := process.NewExternalProcessors()
 
 		alert.Start(config.Alert, config.Pushgw, syncStats, alertStats, externalProcessors, targetCache, busiGroupCache, alertMuteCache,
-			alertRuleCache, notifyConfigCache, taskTplsCache, dsCache, ctx, promClients, tdengineClients, userCache, userGroupCache)
+			alertRuleCache, notifyConfigCache, taskTplsCache, dsCache, ctx, promClients, userCache, userGroupCache)
 
 		alertrtRouter := alertrt.New(config.HTTP, config.Alert, alertMuteCache, targetCache, busiGroupCache, alertStats, ctx, externalProcessors)
 

cron/clean_notify_record.go

@@ -0,0 +1,41 @@
+package cron
+
+import (
+	"time"
+
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
+
+	"github.com/robfig/cron/v3"
+	"github.com/toolkits/pkg/logger"
+)
+
+func cleanNotifyRecord(ctx *ctx.Context, day int) {
+	lastWeek := time.Now().Unix() - 86400*int64(day)
+	err := models.DB(ctx).Model(&models.NotificaitonRecord{}).Where("created_at < ?", lastWeek).Delete(&models.NotificaitonRecord{}).Error
+	if err != nil {
+		logger.Errorf("Failed to clean notify record: %v", err)
+	}
+
+}
+
+// 每天凌晨1点执行清理任务
+func CleanNotifyRecord(ctx *ctx.Context, day int) {
+	c := cron.New()
+	if day < 1 {
+		day = 7
+	}
+
+	// 使用cron表达式设置每天凌晨1点执行
+	_, err := c.AddFunc("0 1 * * *", func() {
+		cleanNotifyRecord(ctx, day)
+	})
+
+	if err != nil {
+		logger.Errorf("Failed to add clean notify record cron job: %v", err)
+		return
+	}
+
+	// 启动cron任务
+	c.Start()
+}

datasource/ck/clickhouse.go

@@ -0,0 +1,241 @@
+package ck
+
+import (
+	"context"
+	"fmt"
+	"strings"
+
+	"github.com/ccfos/nightingale/v6/datasource"
+	ck "github.com/ccfos/nightingale/v6/dskit/clickhouse"
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/macros"
+
+	"github.com/mitchellh/mapstructure"
+	"github.com/toolkits/pkg/logger"
+)
+
+const (
+	CKType = "ck"
+
+	TimeFieldFormatEpochMilli  = "epoch_millis"
+	TimeFieldFormatEpochSecond = "epoch_second"
+
+	DefaultLimit = 500
+)
+
+var (
+	ckPrivBanned = []string{
+		"INSERT",
+		"CREATE",
+		"DROP",
+		"DELETE",
+		"UPDATE",
+		"ALL",
+	}
+
+	ckBannedOp = map[string]struct{}{
+		"CREATE":   {},
+		"INSERT":   {},
+		"ALTER":    {},
+		"REVOKE":   {},
+		"DROP":     {},
+		"RENAME":   {},
+		"ATTACH":   {},
+		"DETACH":   {},
+		"OPTIMIZE": {},
+		"TRUNCATE": {},
+		"SET":      {},
+	}
+)
+
+func init() {
+	datasource.RegisterDatasource(CKType, new(Clickhouse))
+}
+
+type CKShard struct {
+	Addr        string `json:"ck.addr" mapstructure:"ck.addr"`
+	User        string `json:"ck.user" mapstructure:"ck.user"`
+	Password    string `json:"ck.password" mapstructure:"ck.password"`
+	Database    string `json:"ck.db" mapstructure:"ck.db"`
+	IsEncrypted bool   `json:"ck.is_encrypt" mapstructure:"ck.is_encrypt"`
+}
+
+type QueryParam struct {
+	Limit      int             `json:"limit" mapstructure:"limit"`
+	Sql        string          `json:"sql" mapstructure:"sql"`
+	Ref        string          `json:"ref" mapstructure:"ref"`
+	From       int64           `json:"from" mapstructure:"from"`
+	To         int64           `json:"to" mapstructure:"to"`
+	TimeField  string          `json:"time_field" mapstructure:"time_field"`
+	TimeFormat string          `json:"time_format" mapstructure:"time_format"`
+	Keys       datasource.Keys `json:"keys" mapstructure:"keys"`
+	Database   string          `json:"database" mapstructure:"database"`
+	Table      string          `json:"table" mapstructure:"table"`
+}
+
+type Clickhouse struct {
+	ck.Clickhouse `json:",inline" mapstructure:",squash"`
+}
+
+func (c *Clickhouse) Init(settings map[string]interface{}) (datasource.Datasource, error) {
+	newest := new(Clickhouse)
+	err := mapstructure.Decode(settings, newest)
+	return newest, err
+}
+
+func (c *Clickhouse) InitClient() error {
+	return c.InitCli()
+}
+
+func (c *Clickhouse) Validate(ctx context.Context) error {
+	if len(c.Nodes) == 0 {
+		return fmt.Errorf("ck shard is invalid, please check datasource setting")
+	}
+
+	addr := c.Nodes[0]
+	if len(strings.Trim(c.User, " ")) == 0 {
+		return fmt.Errorf("ck shard user is invalid, please check datasource setting")
+	}
+
+	if len(strings.Trim(addr, " ")) == 0 {
+		return fmt.Errorf("ck shard addr is invalid, please check datasource setting")
+	}
+
+	// if len(strings.Trim(shard.Password, " ")) == 0 {
+	// 	return fmt.Errorf("ck shard password is empty, please check datasource setting or set password for user")
+	// }
+
+	return nil
+}
+
+// Equal compares whether two objects are the same, used for caching
+func (c *Clickhouse) Equal(p datasource.Datasource) bool {
+
+	plg, ok := p.(*Clickhouse)
+	if !ok {
+		logger.Errorf("unexpected plugin type, expected is ck")
+		return false
+	}
+
+	// only compare first shard
+	if len(c.Nodes) == 0 {
+		logger.Errorf("ck shard is empty")
+		return false
+	}
+	addr := c.Nodes[0]
+
+	if len(plg.Nodes) == 0 {
+		logger.Errorf("new ck plugin obj shard is empty")
+		return false
+	}
+	newAddr := plg.Nodes[0]
+
+	if c.User != plg.User {
+		return false
+	}
+
+	if addr != newAddr {
+		return false
+	}
+
+	if c.Password != plg.Password {
+		return false
+	}
+
+	return true
+}
+
+func (c *Clickhouse) MakeLogQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	return nil, nil
+}
+
+func (c *Clickhouse) MakeTSQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	return nil, nil
+}
+
+func (c *Clickhouse) QueryMapData(ctx context.Context, query interface{}) ([]map[string]string, error) {
+	return nil, nil
+}
+
+func (c *Clickhouse) QueryData(ctx context.Context, query interface{}) ([]models.DataResp, error) {
+
+	ckQueryParam := new(ck.QueryParam)
+	if err := mapstructure.Decode(query, ckQueryParam); err != nil {
+		return nil, err
+	}
+
+	if strings.Contains(ckQueryParam.Sql, "$__") {
+		var err error
+		ckQueryParam.Sql, err = macros.Macro(ckQueryParam.Sql, ckQueryParam.From, ckQueryParam.To)
+		if err != nil {
+			return nil, err
+		}
+	}
+
+	if ckQueryParam.Keys.ValueKey == "" {
+		return nil, fmt.Errorf("valueKey is required")
+	}
+
+	rows, err := c.QueryTimeseries(ctx, ckQueryParam)
+	if err != nil {
+		logger.Warningf("query:%+v get data err:%v", ckQueryParam, err)
+		return nil, err
+	}
+
+	if err != nil {
+		logger.Warningf("query:%+v get data err:%v", ckQueryParam, err)
+		return []models.DataResp{}, err
+	}
+	data := make([]models.DataResp, 0)
+	for i := range rows {
+		data = append(data, models.DataResp{
+			Ref:    ckQueryParam.Ref,
+			Metric: rows[i].Metric,
+			Values: rows[i].Values,
+		})
+	}
+
+	return data, nil
+}
+
+func (c *Clickhouse) QueryLog(ctx context.Context, query interface{}) ([]interface{}, int64, error) {
+	ckQueryParam := new(QueryParam)
+	if err := mapstructure.Decode(query, ckQueryParam); err != nil {
+		return nil, 0, err
+	}
+
+	if strings.Contains(ckQueryParam.Sql, "$__") {
+		var err error
+		ckQueryParam.Sql, err = macros.Macro(ckQueryParam.Sql, ckQueryParam.From, ckQueryParam.To)
+		if err != nil {
+			return nil, 0, err
+		}
+	}
+
+	rows, err := c.Query(ctx, ckQueryParam)
+	if err != nil {
+		logger.Warningf("query:%+v get data err:%v", ckQueryParam, err)
+		return nil, 0, err
+	}
+
+	limit := getLimit(len(rows), ckQueryParam.Limit)
+
+	logs := make([]interface{}, 0)
+	for i := 0; i < limit; i++ {
+		logs = append(logs, rows[i])
+	}
+
+	return logs, int64(limit), nil
+}
+
+func getLimit(rowLen, pLimit int) int {
+	limit := DefaultLimit
+	if pLimit > 0 {
+		limit = pLimit
+	}
+	if rowLen > limit {
+		return limit
+	}
+
+	return rowLen
+}

datasource/commons/eslike/eslike.go

@@ -0,0 +1,603 @@
+package eslike
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/araddon/dateparse"
+	"github.com/bitly/go-simplejson"
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/mitchellh/mapstructure"
+	"github.com/olivere/elastic/v7"
+	"github.com/prometheus/common/model"
+	"github.com/toolkits/pkg/logger"
+)
+
+type Query struct {
+	Ref        string     `json:"ref" mapstructure:"ref"`
+	Index      string     `json:"index" mapstructure:"index"`
+	Filter     string     `json:"filter" mapstructure:"filter"`
+	MetricAggr MetricAggr `json:"value" mapstructure:"value"`
+	GroupBy    []GroupBy  `json:"group_by" mapstructure:"group_by"`
+	DateField  string     `json:"date_field" mapstructure:"date_field"`
+	Interval   int64      `json:"interval" mapstructure:"interval"`
+	Start      int64      `json:"start" mapstructure:"start"`
+	End        int64      `json:"end" mapstructure:"end"`
+	P          int        `json:"page" mapstructure:"page"`           // 页码
+	Limit      int        `json:"limit" mapstructure:"limit"`         // 每页个数
+	Ascending  bool       `json:"ascending" mapstructure:"ascending"` // 按照DataField排序
+
+	Timeout  int `json:"timeout" mapstructure:"timeout"`
+	MaxShard int `json:"max_shard" mapstructure:"max_shard"`
+}
+
+type MetricAggr struct {
+	Field string `json:"field" mapstructure:"field"`
+	Func  string `json:"func" mapstructure:"func"`
+	Ref   string `json:"ref" mapstructure:"ref"` // 变量名，A B C
+}
+
+type GroupBy struct {
+	Cate        GroupByCate `json:"cate" mapstructure:"cate"` // 分组类型
+	Field       string      `json:"field" mapstructure:"field"`
+	MinDocCount int64       `json:"min_doc_count" mapstructure:"min_doc_count"`
+	Order       string      `json:"order" mapstructure:"order"`
+	OrderBy     string      `json:"order_by" mapstructure:"order_by"`
+	Size        int         `json:"size" mapstructure:"size"`
+
+	Params   []Param `json:"params" mapstructure:"params"`     // 类型是 filter 时使用
+	Interval int64   `json:"interval" mapstructure:"interval"` // 分组间隔
+}
+
+type SearchFunc func(ctx context.Context, indices []string, source interface{}, timeout int, maxShard int) (*elastic.SearchResult, error)
+type QueryFieldsFunc func(indices []string) ([]string, error)
+
+// 分组类型
+type GroupByCate string
+
+const (
+	Filters  GroupByCate = "filters"
+	Histgram GroupByCate = "histgram"
+	Terms    GroupByCate = "terms"
+)
+
+// 参数
+type Param struct {
+	Alias string `json:"alias,omitempty"` // 别名，a=b的形式，filter 特有
+	Query string `json:"query,omitempty"` // 查询条件，filter 特有
+}
+
+type MetricPtr struct {
+	Data map[string][][]float64
+}
+
+func IterGetMap(m, ret map[string]interface{}, prefixKey string) {
+	for k, v := range m {
+		switch v.(type) {
+		case map[string]interface{}:
+			var key string
+			if prefixKey != "" {
+				key = fmt.Sprintf("%s.%s", prefixKey, k)
+			} else {
+				key = k
+			}
+			IterGetMap(v.(map[string]interface{}), ret, key)
+		default:
+			ret[prefixKey+"."+k] = []interface{}{v}
+		}
+	}
+}
+
+func TransferData(metric, ref string, m map[string][][]float64) []models.DataResp {
+	var datas []models.DataResp
+
+	for k, v := range m {
+		data := models.DataResp{
+			Ref:    ref,
+			Metric: make(model.Metric),
+			Labels: k,
+			Values: v,
+		}
+
+		data.Metric["__name__"] = model.LabelValue(metric)
+		labels := strings.Split(k, "--")
+		for _, label := range labels {
+			arr := strings.SplitN(label, "=", 2)
+			if len(arr) == 2 {
+				data.Metric[model.LabelName(arr[0])] = model.LabelValue(arr[1])
+			}
+		}
+		datas = append(datas, data)
+	}
+
+	for i := 0; i < len(datas); i++ {
+		for k, v := range datas[i].Metric {
+			if k == "__name__" {
+				datas[i].Metric[k] = model.LabelValue(ref) + "_" + v
+			}
+		}
+	}
+
+	return datas
+}
+
+func GetQueryString(filter string, q *elastic.RangeQuery) *elastic.BoolQuery {
+	var queryString *elastic.BoolQuery
+	if filter != "" {
+		if strings.Contains(filter, ":") || strings.Contains(filter, "AND") || strings.Contains(filter, "OR") || strings.Contains(filter, "NOT") {
+			queryString = elastic.NewBoolQuery().Must(elastic.NewQueryStringQuery(filter)).Filter(q)
+		} else {
+			queryString = elastic.NewBoolQuery().Filter(elastic.NewMultiMatchQuery(filter).Lenient(true).Type("phrase")).Filter(q)
+		}
+	} else {
+		queryString = elastic.NewBoolQuery().Should(q)
+	}
+
+	return queryString
+}
+
+func getUnixTs(timeStr string) int64 {
+	ts, err := strconv.ParseInt(timeStr, 10, 64)
+	if err == nil {
+		return ts
+	}
+
+	parsedTime, err := dateparse.ParseAny(timeStr)
+	if err != nil {
+		logger.Error("failed to ParseAny: ", err)
+		return 0
+	}
+	return parsedTime.UnixMilli()
+}
+
+func GetBuckts(labelKey string, keys []string, arr []interface{}, metrics *MetricPtr, labels string, ts int64, f string) {
+	var err error
+	bucketsKey := ""
+	if len(keys) > 0 {
+		bucketsKey = keys[0]
+	}
+
+	newlabels := ""
+	for i := 0; i < len(arr); i++ {
+		tmp := arr[i].(map[string]interface{})
+		keyAsString, getTs := tmp["key_as_string"]
+		if getTs {
+			ts = getUnixTs(keyAsString.(string))
+		}
+		keyValue := tmp["key"]
+		switch keyValue.(type) {
+		case json.Number, string:
+			if !getTs {
+				if labels != "" {
+					newlabels = fmt.Sprintf("%s--%s=%v", labels, labelKey, keyValue)
+				} else {
+					newlabels = fmt.Sprintf("%s=%v", labelKey, keyValue)
+				}
+			}
+		default:
+			continue
+		}
+
+		var finalValue float64
+		if len(keys) == 0 { // 计算 doc_count 的情况
+			count := tmp["doc_count"]
+			finalValue, err = count.(json.Number).Float64()
+			if err != nil {
+				logger.Warningf("labelKey:%s get value error:%v", labelKey, err)
+			}
+			newValues := []float64{float64(ts / 1000), finalValue}
+			metrics.Data[newlabels] = append(metrics.Data[newlabels], newValues)
+			continue
+		}
+
+		innerBuckets, exists := tmp[bucketsKey]
+		if !exists {
+			continue
+		}
+
+		nextBucketsArr, exists := innerBuckets.(map[string]interface{})["buckets"]
+		if exists {
+			if len(keys[1:]) >= 1 {
+				GetBuckts(bucketsKey, keys[1:], nextBucketsArr.([]interface{}), metrics, newlabels, ts, f)
+			} else {
+				GetBuckts(bucketsKey, []string{}, nextBucketsArr.([]interface{}), metrics, newlabels, ts, f)
+			}
+		} else {
+
+			// doc_count
+			if f == "count" || f == "nodata" {
+				count := tmp["doc_count"]
+				finalValue, err = count.(json.Number).Float64()
+				if err != nil {
+					logger.Warningf("get %v value error:%v", count, err)
+				}
+			} else {
+				values, exists := innerBuckets.(map[string]interface{})["value"]
+				if exists {
+					switch values.(type) {
+					case json.Number:
+						value, err := values.(json.Number).Float64()
+						if err != nil {
+							logger.Warningf("labelKey:%s get value error:%v", labelKey, err)
+						}
+						finalValue = value
+					}
+				} else {
+					switch values.(type) {
+					case map[string]interface{}:
+						var err error
+						values := innerBuckets.(map[string]interface{})["values"]
+						for _, v := range values.(map[string]interface{}) {
+							finalValue, err = v.(json.Number).Float64()
+							if err != nil {
+								logger.Warningf("labelKey:%s get value error:%v", labelKey, err)
+							}
+						}
+					default:
+						values := innerBuckets.(map[string]interface{})["values"]
+						for _, v := range values.(map[string]interface{}) {
+							// Todo 修复 v is nil 导致 panic 情况
+							finalValue, err = v.(json.Number).Float64()
+							if err != nil {
+								logger.Warningf("labelKey:%s get value error:%v", labelKey, err)
+							}
+						}
+					}
+				}
+			}
+
+			if _, exists := metrics.Data[newlabels]; !exists {
+				metrics.Data[newlabels] = [][]float64{}
+			}
+
+			newValues := []float64{float64(ts / 1000), finalValue}
+			metrics.Data[newlabels] = append(metrics.Data[newlabels], newValues)
+		}
+	}
+}
+
+func MakeLogQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	param := new(Query)
+	if err := mapstructure.Decode(query, param); err != nil {
+		return nil, err
+	}
+
+	for i := 0; i < len(eventTags); i++ {
+		eventTags[i] = strings.Replace(eventTags[i], "=", ":", 1)
+	}
+
+	if len(eventTags) > 0 {
+		if param.Filter == "" {
+			param.Filter = strings.Join(eventTags, " AND ")
+		} else {
+			param.Filter = param.Filter + " AND " + strings.Join(eventTags, " AND ")
+		}
+	}
+
+	param.Start = start
+	param.End = end
+
+	return param, nil
+}
+
+func MakeTSQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	param := new(Query)
+	if err := mapstructure.Decode(query, param); err != nil {
+		return nil, err
+	}
+
+	for i := 0; i < len(eventTags); i++ {
+		eventTags[i] = strings.Replace(eventTags[i], "=", ":", 1)
+	}
+
+	if len(eventTags) > 0 {
+		if param.Filter == "" {
+			param.Filter = strings.Join(eventTags, " AND ")
+		} else {
+			param.Filter = param.Filter + " AND " + strings.Join(eventTags, " AND ")
+		}
+	}
+	param.Start = start
+	param.End = end
+
+	return param, nil
+}
+
+func QueryData(ctx context.Context, queryParam interface{}, cliTimeout int64, version string, search SearchFunc) ([]models.DataResp, error) {
+	param := new(Query)
+	if err := mapstructure.Decode(queryParam, param); err != nil {
+		return nil, err
+	}
+
+	if param.Timeout == 0 {
+		param.Timeout = int(cliTimeout) / 1000
+	}
+
+	if param.Interval == 0 {
+		param.Interval = 60
+	}
+
+	if param.MaxShard < 1 {
+		param.MaxShard = 5
+	}
+
+	if param.DateField == "" {
+		param.DateField = "@timestamp"
+	}
+
+	indexArr := strings.Split(param.Index, ",")
+	q := elastic.NewRangeQuery(param.DateField)
+	now := time.Now().Unix()
+	var start, end int64
+	if param.End != 0 && param.Start != 0 {
+		end = param.End - param.End%param.Interval
+		start = param.Start - param.Start%param.Interval
+	} else {
+		end = now
+		start = end - param.Interval
+	}
+
+	delay, ok := ctx.Value("delay").(int64)
+	if ok && delay != 0 {
+		end = end - delay
+		start = start - delay
+	}
+
+	q.Gte(time.Unix(start, 0).UnixMilli())
+	q.Lte(time.Unix(end, 0).UnixMilli())
+	q.Format("epoch_millis")
+
+	field := param.MetricAggr.Field
+	groupBys := param.GroupBy
+
+	queryString := GetQueryString(param.Filter, q)
+
+	var aggr elastic.Aggregation
+	switch param.MetricAggr.Func {
+	case "avg":
+		aggr = elastic.NewAvgAggregation().Field(field)
+	case "max":
+		aggr = elastic.NewMaxAggregation().Field(field)
+	case "min":
+		aggr = elastic.NewMinAggregation().Field(field)
+	case "sum":
+		aggr = elastic.NewSumAggregation().Field(field)
+	case "count":
+		aggr = elastic.NewValueCountAggregation().Field(field)
+	case "p90":
+		aggr = elastic.NewPercentilesAggregation().Percentiles(90).Field(field)
+	case "p95":
+		aggr = elastic.NewPercentilesAggregation().Percentiles(95).Field(field)
+	case "p99":
+		aggr = elastic.NewPercentilesAggregation().Percentiles(99).Field(field)
+	case "median":
+		aggr = elastic.NewPercentilesAggregation().Percentiles(50).Field(field)
+	default:
+		return nil, fmt.Errorf("func %s not support", param.MetricAggr.Func)
+	}
+
+	tsAggr := elastic.NewDateHistogramAggregation().
+		Field(param.DateField).
+		MinDocCount(1)
+
+	if strings.HasPrefix(version, "7") {
+		tsAggr.FixedInterval(fmt.Sprintf("%ds", param.Interval))
+	} else {
+		// 兼容 7.0 以下的版本
+		// OpenSearch 也使用这个字段
+		tsAggr.Interval(fmt.Sprintf("%ds", param.Interval))
+	}
+
+	// group by
+	var groupByAggregation elastic.Aggregation
+	if len(groupBys) > 0 {
+		groupBy := groupBys[0]
+
+		if groupBy.MinDocCount == 0 {
+			groupBy.MinDocCount = 1
+		}
+
+		if groupBy.Size == 0 {
+			groupBy.Size = 300
+		}
+
+		switch groupBy.Cate {
+		case Terms:
+			if param.MetricAggr.Func != "count" {
+				groupByAggregation = elastic.NewTermsAggregation().Field(groupBy.Field).SubAggregation(field, aggr).OrderByKeyDesc().Size(groupBy.Size).MinDocCount(int(groupBy.MinDocCount))
+			} else {
+				groupByAggregation = elastic.NewTermsAggregation().Field(groupBy.Field).OrderByKeyDesc().Size(groupBy.Size).MinDocCount(int(groupBy.MinDocCount))
+			}
+		case Histgram:
+			if param.MetricAggr.Func != "count" {
+				groupByAggregation = elastic.NewHistogramAggregation().Field(groupBy.Field).Interval(float64(groupBy.Interval)).SubAggregation(field, aggr)
+			} else {
+				groupByAggregation = elastic.NewHistogramAggregation().Field(groupBy.Field).Interval(float64(groupBy.Interval))
+			}
+		case Filters:
+			for _, filterParam := range groupBy.Params {
+				if param.MetricAggr.Func != "count" {
+					groupByAggregation = elastic.NewFilterAggregation().Filter(elastic.NewTermQuery(filterParam.Query, "true")).SubAggregation(field, aggr)
+				} else {
+					groupByAggregation = elastic.NewFilterAggregation().Filter(elastic.NewTermQuery(filterParam.Query, "true"))
+				}
+			}
+		}
+
+		for i := 1; i < len(groupBys); i++ {
+			groupBy := groupBys[i]
+
+			if groupBy.MinDocCount == 0 {
+				groupBy.MinDocCount = 1
+			}
+
+			if groupBy.Size == 0 {
+				groupBy.Size = 300
+			}
+
+			switch groupBy.Cate {
+			case Terms:
+				groupByAggregation = elastic.NewTermsAggregation().Field(groupBy.Field).SubAggregation(groupBys[i-1].Field, groupByAggregation).OrderByKeyDesc().Size(groupBy.Size).MinDocCount(int(groupBy.MinDocCount))
+			case Histgram:
+				groupByAggregation = elastic.NewHistogramAggregation().Field(groupBy.Field).Interval(float64(groupBy.Interval)).SubAggregation(groupBys[i-1].Field, groupByAggregation)
+			case Filters:
+				for _, filterParam := range groupBy.Params {
+					groupByAggregation = elastic.NewFilterAggregation().Filter(elastic.NewTermQuery(filterParam.Query, "true")).SubAggregation(groupBys[i-1].Field, groupByAggregation)
+				}
+			}
+		}
+
+		tsAggr.SubAggregation(groupBys[len(groupBys)-1].Field, groupByAggregation)
+	} else if param.MetricAggr.Func != "count" {
+		tsAggr.SubAggregation(field, aggr)
+	}
+
+	source, _ := queryString.Source()
+	b, _ := json.Marshal(source)
+	logger.Debugf("query_data q:%+v tsAggr:%+v query_string:%s", param, tsAggr, string(b))
+
+	searchSource := elastic.NewSearchSource().
+		Query(queryString).
+		Aggregation("ts", tsAggr)
+
+	searchSourceString, err := searchSource.Source()
+	if err != nil {
+		logger.Warningf("query_data searchSource:%s to string error:%v", searchSourceString, err)
+	}
+
+	jsonSearchSource, err := json.Marshal(searchSourceString)
+	if err != nil {
+		logger.Warningf("query_data searchSource:%s to json error:%v", searchSourceString, err)
+	}
+
+	result, err := search(ctx, indexArr, searchSource, param.Timeout, param.MaxShard)
+	if err != nil {
+		logger.Warningf("query_data searchSource:%s query_data error:%v", searchSourceString, err)
+		return nil, err
+	}
+
+	logger.Debugf("query_data searchSource:%s resp:%s", string(jsonSearchSource), string(result.Aggregations["ts"]))
+
+	js, err := simplejson.NewJson(result.Aggregations["ts"])
+	if err != nil {
+		return nil, err
+	}
+
+	bucketsData, err := js.Get("buckets").Array()
+	if err != nil {
+		return nil, err
+	}
+
+	var keys []string
+	for i := len(groupBys) - 1; i >= 0; i-- {
+		keys = append(keys, groupBys[i].Field)
+	}
+
+	if param.MetricAggr.Func != "count" {
+		keys = append(keys, field)
+	}
+
+	metrics := &MetricPtr{Data: make(map[string][][]float64)}
+
+	GetBuckts("", keys, bucketsData, metrics, "", 0, param.MetricAggr.Func)
+
+	return TransferData(fmt.Sprintf("%s_%s", field, param.MetricAggr.Func), param.Ref, metrics.Data), nil
+}
+
+func HitFilter(typ string) bool {
+	switch typ {
+	case "keyword", "date", "long", "integer", "short", "byte", "double", "float", "half_float", "scaled_float", "unsigned_long":
+		return false
+	default:
+		return true
+	}
+}
+
+func QueryLog(ctx context.Context, queryParam interface{}, timeout int64, version string, maxShard int, search SearchFunc) ([]interface{}, int64, error) {
+	param := new(Query)
+	if err := mapstructure.Decode(queryParam, param); err != nil {
+		return nil, 0, err
+	}
+
+	if param.Timeout == 0 {
+		param.Timeout = int(timeout)
+	}
+
+	indexArr := strings.Split(param.Index, ",")
+
+	now := time.Now().Unix()
+	var start, end int64
+	if param.End != 0 && param.Start != 0 {
+		end = param.End - param.End%param.Interval
+		start = param.Start - param.Start%param.Interval
+	} else {
+		end = now
+		start = end - param.Interval
+	}
+
+	q := elastic.NewRangeQuery(param.DateField)
+	q.Gte(time.Unix(start, 0).UnixMilli())
+	q.Lte(time.Unix(end, 0).UnixMilli())
+	q.Format("epoch_millis")
+
+	queryString := GetQueryString(param.Filter, q)
+
+	if param.Limit <= 0 {
+		param.Limit = 10
+	}
+
+	if param.MaxShard < 1 {
+		param.MaxShard = maxShard
+	}
+
+	source := elastic.NewSearchSource().
+		TrackTotalHits(true).
+		Query(queryString).
+		From(param.P).
+		Size(param.Limit).
+		Sort(param.DateField, param.Ascending)
+
+	result, err := search(ctx, indexArr, source, param.Timeout, param.MaxShard)
+	if err != nil {
+		logger.Warningf("query data error:%v", err)
+		return nil, 0, err
+	}
+
+	total := result.TotalHits()
+
+	var ret []interface{}
+
+	b, _ := json.Marshal(source)
+	logger.Debugf("query data result query source:%s len:%d total:%d", string(b), len(result.Hits.Hits), total)
+
+	resultBytes, _ := json.Marshal(result)
+	logger.Debugf("query data result query source:%s result:%s", string(b), string(resultBytes))
+
+	if strings.HasPrefix(version, "6") {
+		for i := 0; i < len(result.Hits.Hits); i++ {
+			var x map[string]interface{}
+			err := json.Unmarshal(result.Hits.Hits[i].Source, &x)
+			if err != nil {
+				logger.Warningf("Unmarshal soruce error:%v", err)
+				continue
+			}
+
+			if result.Hits.Hits[i].Fields == nil {
+				result.Hits.Hits[i].Fields = make(map[string]interface{})
+			}
+
+			IterGetMap(x, result.Hits.Hits[i].Fields, "")
+			ret = append(ret, result.Hits.Hits[i])
+		}
+	} else {
+		for _, hit := range result.Hits.Hits {
+			ret = append(ret, hit)
+		}
+	}
+
+	return ret, total, nil
+}

datasource/datasource.go

@@ -0,0 +1,114 @@
+package datasource
+
+import (
+	"context"
+	"fmt"
+	"strings"
+
+	"github.com/ccfos/nightingale/v6/models"
+)
+
+type DatasourceType struct {
+	Id             int64  `json:"id"`
+	Category       string `json:"category"`
+	PluginType     string `json:"type"`
+	PluginTypeName string `json:"type_name"`
+}
+
+type Keys struct {
+	ValueKey   string `json:"valueKey" mapstructure:"valueKey"` // 多个用空格分隔
+	LabelKey   string `json:"labelKey" mapstructure:"labelKey"` // 多个用空格分隔
+	TimeKey    string `json:"timeKey" mapstructure:"timeKey"`
+	TimeFormat string `json:"timeFormat" mapstructure:"timeFormat"`
+}
+
+var DatasourceTypes map[int64]DatasourceType
+
+func init() {
+	DatasourceTypes = make(map[int64]DatasourceType)
+	DatasourceTypes[1] = DatasourceType{
+		Id:             1,
+		Category:       "timeseries",
+		PluginType:     "prometheus",
+		PluginTypeName: "Prometheus Like",
+	}
+
+	DatasourceTypes[2] = DatasourceType{
+		Id:             2,
+		Category:       "logging",
+		PluginType:     "elasticsearch",
+		PluginTypeName: "Elasticsearch",
+	}
+
+	DatasourceTypes[3] = DatasourceType{
+		Id:             3,
+		Category:       "logging",
+		PluginType:     "aliyun-sls",
+		PluginTypeName: "SLS",
+	}
+
+	DatasourceTypes[4] = DatasourceType{
+		Id:             4,
+		Category:       "timeseries",
+		PluginType:     "ck",
+		PluginTypeName: "ClickHouse",
+	}
+}
+
+type NewDatasrouceFn func(settings map[string]interface{}) (Datasource, error)
+
+var datasourceRegister = map[string]NewDatasrouceFn{}
+
+type Datasource interface {
+	Init(settings map[string]interface{}) (Datasource, error) // 初始化配置
+	InitClient() error                                        // 初始化客户端
+	Validate(ctx context.Context) error                       // 参数验证
+	Equal(p Datasource) bool                                  // 验证是否相等
+	MakeLogQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error)
+	MakeTSQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error)
+	QueryData(ctx context.Context, query interface{}) ([]models.DataResp, error)
+	QueryLog(ctx context.Context, query interface{}) ([]interface{}, int64, error)
+
+	// 在生成告警事件时，会调用该方法，用于获取额外的数据
+	QueryMapData(ctx context.Context, query interface{}) ([]map[string]string, error)
+}
+
+func RegisterDatasource(typ string, p Datasource) {
+	if _, found := datasourceRegister[typ]; found {
+		return
+	}
+	datasourceRegister[typ] = p.Init
+}
+
+func GetDatasourceByType(typ string, settings map[string]interface{}) (Datasource, error) {
+	typ = strings.ReplaceAll(typ, ".logging", "")
+	fn, found := datasourceRegister[typ]
+	if !found {
+		return nil, fmt.Errorf("plugin type %s not found", typ)
+	}
+
+	plug, err := fn(settings)
+	if err != nil {
+		return nil, err
+	}
+
+	return plug, nil
+}
+
+type DatasourceInfo struct {
+	Id             int64                  `json:"id"`
+	Name           string                 `json:"name"`
+	Description    string                 `json:"description"`
+	ClusterName    string                 `json:"cluster_name"`
+	Category       string                 `json:"category"`
+	PluginId       int64                  `json:"plugin_id"`
+	Type           string                 `json:"plugin_type"`
+	PluginTypeName string                 `json:"plugin_type_name"`
+	Settings       map[string]interface{} `json:"settings"`
+	HTTPJson       models.HTTP            `json:"http"`
+	AuthJson       models.Auth            `json:"auth"`
+	Status         string                 `json:"status"`
+	CreatedAt      int64                  `json:"created_at"`
+	UpdatedAt      int64                  `json:"updated_at"`
+	IsDefault      bool                   `json:"is_default"`
+}

datasource/es/es.go

@@ -0,0 +1,405 @@
+package es
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"net"
+	"net/http"
+	"net/url"
+	"reflect"
+	"sort"
+	"strings"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/datasource"
+	"github.com/ccfos/nightingale/v6/datasource/commons/eslike"
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/tlsx"
+
+	"github.com/mitchellh/mapstructure"
+	"github.com/olivere/elastic/v7"
+	"github.com/toolkits/pkg/logger"
+)
+
+const (
+	ESType = "elasticsearch"
+)
+
+type Elasticsearch struct {
+	Addr        string            `json:"es.addr" mapstructure:"es.addr"`
+	Nodes       []string          `json:"es.nodes" mapstructure:"es.nodes"`
+	Timeout     int64             `json:"es.timeout" mapstructure:"es.timeout"` // millis
+	Basic       BasicAuth         `json:"es.basic" mapstructure:"es.basic"`
+	TLS         TLS               `json:"es.tls" mapstructure:"es.tls"`
+	Version     string            `json:"es.version" mapstructure:"es.version"`
+	Headers     map[string]string `json:"es.headers" mapstructure:"es.headers"`
+	MinInterval int               `json:"es.min_interval" mapstructure:"es.min_interval"` // seconds
+	MaxShard    int               `json:"es.max_shard" mapstructure:"es.max_shard"`
+	ClusterName string            `json:"es.cluster_name" mapstructure:"es.cluster_name"`
+	EnableWrite bool              `json:"es.enable_write" mapstructure:"es.enable_write"` // 允许写操作
+	Client      *elastic.Client   `json:"es.client" mapstructure:"es.client"`
+}
+
+type TLS struct {
+	SkipTlsVerify bool `json:"es.tls.skip_tls_verify" mapstructure:"es.tls.skip_tls_verify"`
+}
+
+type BasicAuth struct {
+	Enable   bool   `json:"es.auth.enable" mapstructure:"es.auth.enable"`
+	Username string `json:"es.user" mapstructure:"es.user"`
+	Password string `json:"es.password" mapstructure:"es.password"`
+}
+
+func init() {
+	datasource.RegisterDatasource(ESType, new(Elasticsearch))
+}
+
+func (e *Elasticsearch) Init(settings map[string]interface{}) (datasource.Datasource, error) {
+	newest := new(Elasticsearch)
+	err := mapstructure.Decode(settings, newest)
+	return newest, err
+}
+
+func (e *Elasticsearch) InitClient() error {
+	transport := &http.Transport{
+		Proxy: http.ProxyFromEnvironment,
+		DialContext: (&net.Dialer{
+			Timeout: time.Duration(e.Timeout) * time.Millisecond,
+		}).DialContext,
+		ResponseHeaderTimeout: time.Duration(e.Timeout) * time.Millisecond,
+	}
+
+	if len(e.Nodes) > 0 {
+		e.Addr = e.Nodes[0]
+	}
+
+	if strings.Contains(e.Addr, "https") {
+		tlsConfig := tlsx.ClientConfig{
+			InsecureSkipVerify: e.TLS.SkipTlsVerify,
+			UseTLS:             true,
+		}
+		cfg, err := tlsConfig.TLSConfig()
+		if err != nil {
+			return err
+		}
+		transport.TLSClientConfig = cfg
+	}
+
+	var err error
+	options := []elastic.ClientOptionFunc{
+		elastic.SetURL(e.Nodes...),
+	}
+
+	if e.Basic.Username != "" {
+		options = append(options, elastic.SetBasicAuth(e.Basic.Username, e.Basic.Password))
+	}
+
+	headers := http.Header{}
+	for k, v := range e.Headers {
+		headers[k] = []string{v}
+	}
+
+	options = append(options, elastic.SetHeaders(headers))
+	options = append(options, elastic.SetHttpClient(&http.Client{Transport: transport}))
+	options = append(options, elastic.SetSniff(false))
+	options = append(options, elastic.SetHealthcheck(false))
+
+	e.Client, err = elastic.NewClient(options...)
+	return err
+}
+
+func (e *Elasticsearch) Equal(other datasource.Datasource) bool {
+	sort.Strings(e.Nodes)
+	sort.Strings(other.(*Elasticsearch).Nodes)
+
+	if strings.Join(e.Nodes, ",") != strings.Join(other.(*Elasticsearch).Nodes, ",") {
+		return false
+	}
+
+	if e.Basic.Username != other.(*Elasticsearch).Basic.Username {
+		return false
+	}
+
+	if e.Basic.Password != other.(*Elasticsearch).Basic.Password {
+		return false
+	}
+
+	if e.TLS.SkipTlsVerify != other.(*Elasticsearch).TLS.SkipTlsVerify {
+		return false
+	}
+	if e.EnableWrite != other.(*Elasticsearch).EnableWrite {
+		return false
+	}
+
+	if !reflect.DeepEqual(e.Headers, other.(*Elasticsearch).Headers) {
+		return false
+	}
+
+	return true
+}
+
+func (e *Elasticsearch) Validate(ctx context.Context) (err error) {
+	if len(e.Nodes) == 0 {
+		return fmt.Errorf("need a valid addr")
+	}
+
+	for _, addr := range e.Nodes {
+		_, err = url.Parse(addr)
+		if err != nil {
+			return fmt.Errorf("parse addr error: %v", err)
+		}
+	}
+
+	if e.Basic.Enable && (len(e.Basic.Username) == 0 || len(e.Basic.Password) == 0) {
+		return fmt.Errorf("need a valid user, password")
+	}
+
+	if e.MaxShard == 0 {
+		e.MaxShard = 5
+	}
+
+	if e.MinInterval < 10 {
+		e.MinInterval = 10
+	}
+
+	if e.Timeout == 0 {
+		e.Timeout = 60000
+	}
+
+	if !strings.HasPrefix(e.Version, "6") && !strings.HasPrefix(e.Version, "7") {
+		return fmt.Errorf("version must be 6.0+ or 7.0+")
+	}
+
+	return nil
+}
+
+func (e *Elasticsearch) MakeLogQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	return eslike.MakeLogQuery(ctx, query, eventTags, start, end)
+}
+
+func (e *Elasticsearch) MakeTSQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	return eslike.MakeTSQuery(ctx, query, eventTags, start, end)
+}
+
+func (e *Elasticsearch) QueryData(ctx context.Context, queryParam interface{}) ([]models.DataResp, error) {
+
+	search := func(ctx context.Context, indices []string, source interface{}, timeout int, maxShard int) (*elastic.SearchResult, error) {
+		return e.Client.Search().
+			Index(indices...).
+			Source(source).
+			Timeout(fmt.Sprintf("%ds", timeout)).
+			MaxConcurrentShardRequests(maxShard).
+			Do(ctx)
+	}
+
+	return eslike.QueryData(ctx, queryParam, e.Timeout, e.Version, search)
+}
+
+func (e *Elasticsearch) QueryIndices() ([]string, error) {
+	result, err := e.Client.IndexNames()
+
+	return result, err
+}
+
+func (e *Elasticsearch) QueryFields(indexs []string) ([]string, error) {
+	var fields []string
+	result, err := elastic.NewGetFieldMappingService(e.Client).Index(indexs...).Do(context.Background())
+	if err != nil {
+		return fields, err
+	}
+
+	fieldMap := make(map[string]struct{})
+	for _, indexMap := range result {
+		if m, exists := indexMap.(map[string]interface{})["mappings"]; exists {
+			for k, v := range m.(map[string]interface{}) {
+				// 兼容 es6 版本
+				if k == "doc" && strings.HasPrefix(e.Version, "6") {
+					// if k == "doc" {
+					for kk, vv := range v.(map[string]interface{}) {
+						typ := getFieldType(kk, vv.(map[string]interface{}))
+						if eslike.HitFilter(typ) {
+							continue
+						}
+
+						if _, exsits := fieldMap[kk]; !exsits {
+							fieldMap[kk] = struct{}{}
+							fields = append(fields, kk)
+						}
+					}
+				} else {
+					// es7 版本
+					typ := getFieldType(k, v.(map[string]interface{}))
+					if eslike.HitFilter(typ) {
+						continue
+					}
+
+					if _, exsits := fieldMap[k]; !exsits {
+						fieldMap[k] = struct{}{}
+						fields = append(fields, k)
+					}
+				}
+
+			}
+		}
+	}
+
+	sort.Strings(fields)
+	return fields, nil
+}
+
+func (e *Elasticsearch) QueryLog(ctx context.Context, queryParam interface{}) ([]interface{}, int64, error) {
+
+	search := func(ctx context.Context, indices []string, source interface{}, timeout int, maxShard int) (*elastic.SearchResult, error) {
+		// 应该是之前为了获取 fields 字段，做的这个兼容
+		// fields, err := e.QueryFields(indices)
+		// if err != nil {
+		// 	logger.Warningf("query data error:%v", err)
+		// 	return nil, err
+		// }
+
+		// if source != nil && strings.HasPrefix(e.Version, "7") {
+		// 	source = source.(*elastic.SearchSource).DocvalueFields(fields...)
+		// }
+
+		return e.Client.Search().
+			Index(indices...).
+			MaxConcurrentShardRequests(maxShard).
+			Source(source).
+			Timeout(fmt.Sprintf("%ds", timeout)).
+			Do(ctx)
+	}
+
+	return eslike.QueryLog(ctx, queryParam, e.Timeout, e.Version, e.MaxShard, search)
+}
+
+func (e *Elasticsearch) QueryFieldValue(indexs []string, field string, query string) ([]string, error) {
+	var values []string
+	search := e.Client.Search().
+		Index(indexs...).
+		Size(0)
+
+	if query != "" {
+		search = search.Query(elastic.NewBoolQuery().Must(elastic.NewQueryStringQuery(query)))
+	}
+	search = search.Aggregation("distinct", elastic.NewTermsAggregation().Field(field).Size(10000))
+
+	result, err := search.Do(context.Background())
+	if err != nil {
+		return values, err
+	}
+
+	agg, found := result.Aggregations.Terms("distinct")
+	if !found {
+		return values, nil
+	}
+
+	for _, bucket := range agg.Buckets {
+		values = append(values, bucket.Key.(string))
+	}
+
+	return values, nil
+}
+
+func (e *Elasticsearch) Test(ctx context.Context) (err error) {
+	err = e.Validate(ctx)
+	if err != nil {
+		return err
+	}
+
+	if e.Addr == "" {
+		return fmt.Errorf("addr is invalid")
+	}
+
+	if e.Version == "7.10+" {
+		options := []elastic.ClientOptionFunc{
+			elastic.SetURL(e.Addr),
+		}
+
+		if e.Basic.Enable {
+			options = append(options, elastic.SetBasicAuth(e.Basic.Username, e.Basic.Password))
+		}
+
+		client, err := elastic.NewClient(options...)
+		if err != nil {
+			return fmt.Errorf("config is invalid:%v", err)
+		}
+
+		_, err = client.ElasticsearchVersion(e.Addr)
+		if err != nil {
+			return fmt.Errorf("config is invalid:%v", err)
+		}
+	} else {
+		return fmt.Errorf("version must be 7.10+")
+	}
+
+	return nil
+}
+
+func getFieldType(key string, m map[string]interface{}) string {
+	if innerMap, exists := m["mapping"]; exists {
+		if innerM, exists := innerMap.(map[string]interface{})[key]; exists {
+			if typ, exists := innerM.(map[string]interface{})["type"]; exists {
+				return typ.(string)
+			}
+		} else {
+			arr := strings.Split(key, ".")
+			if innerM, exists := innerMap.(map[string]interface{})[arr[len(arr)-1]]; exists {
+				if typ, exists := innerM.(map[string]interface{})["type"]; exists {
+					return typ.(string)
+				}
+			}
+		}
+	}
+	return ""
+}
+
+func (e *Elasticsearch) QueryMapData(ctx context.Context, query interface{}) ([]map[string]string, error) {
+	search := func(ctx context.Context, indices []string, source interface{}, timeout int, maxShard int) (*elastic.SearchResult, error) {
+
+		return e.Client.Search().
+			Index(indices...).
+			Source(source).
+			Timeout(fmt.Sprintf("%ds", timeout)).
+			Do(ctx)
+	}
+
+	param := new(eslike.Query)
+	if err := mapstructure.Decode(query, param); err != nil {
+		return nil, err
+	}
+	// 扩大查询范围, 解决上一次查询消耗时间太多，导致本次查询时间范围起止时间，滞后问题
+	param.Interval += 30
+
+	res, _, err := eslike.QueryLog(ctx, param, e.Timeout, e.Version, e.MaxShard, search)
+	if err != nil {
+		return nil, err
+	}
+
+	var result []map[string]string
+	for _, item := range res {
+		logger.Debugf("query:%v item:%v", query, item)
+		if itemMap, ok := item.(*elastic.SearchHit); ok {
+			mItem := make(map[string]string)
+			// 遍历 fields 字段的每个键值对
+			sourceMap := make(map[string]interface{})
+			err := json.Unmarshal(itemMap.Source, &sourceMap)
+			if err != nil {
+				logger.Warningf("unmarshal source%s error:%v", string(itemMap.Source), err)
+				continue
+			}
+
+			for k, v := range sourceMap {
+				mItem[k] = fmt.Sprintf("%v", v)
+			}
+
+			// 将处理好的 map 添加到 m 切片中
+			result = append(result, mItem)
+
+			// 只取第一条数据
+			break
+		}
+	}
+
+	return result, nil
+}

datasource/prom/prom.go

@@ -0,0 +1,15 @@
+package prom
+
+type Prometheus struct {
+	PrometheusAddr  string `json:"prometheus.addr"`
+	PrometheusBasic struct {
+		PrometheusUser string `json:"prometheus.user"`
+		PrometheusPass string `json:"prometheus.password"`
+	} `json:"prometheus.basic"`
+	Headers           map[string]string `json:"prometheus.headers"`
+	PrometheusTimeout int64             `json:"prometheus.timeout"`
+	ClusterName       string            `json:"prometheus.cluster_name"`
+	WriteAddr         string            `json:"prometheus.write_addr"`
+	TsdbType          string            `json:"prometheus.tsdb_type"`
+	InternalAddr      string            `json:"prometheus.internal_addr"`
+}

datasource/tdengine/tdengine.go

@@ -0,0 +1,460 @@
+package tdengine
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"reflect"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/prometheus/common/model"
+	"github.com/toolkits/pkg/logger"
+
+	"github.com/ccfos/nightingale/v6/datasource"
+	td "github.com/ccfos/nightingale/v6/dskit/tdengine"
+	"github.com/ccfos/nightingale/v6/models"
+
+	"github.com/mitchellh/mapstructure"
+)
+
+const (
+	TDEngineType = "tdengine"
+)
+
+type TDengine struct {
+	td.Tdengine `json:",inline" mapstructure:",squash"`
+}
+
+type TdengineQuery struct {
+	From     string `json:"from"`
+	Interval int64  `json:"interval"`
+	Keys     Keys   `json:"keys"`
+	Query    string `json:"query"` // 查询条件
+	Ref      string `json:"ref"`   // 变量
+	To       string `json:"to"`
+}
+
+type Keys struct {
+	LabelKey   string `json:"labelKey"`  // 多个用空格分隔
+	MetricKey  string `json:"metricKey"` // 多个用空格分隔
+	TimeFormat string `json:"timeFormat"`
+}
+
+func init() {
+	datasource.RegisterDatasource(TDEngineType, new(TDengine))
+}
+
+func (td *TDengine) Init(settings map[string]interface{}) (datasource.Datasource, error) {
+	newest := new(TDengine)
+	err := mapstructure.Decode(settings, newest)
+	return newest, err
+}
+
+func (td *TDengine) InitClient() error {
+	td.InitCli()
+	return nil
+}
+
+func (td *TDengine) Equal(other datasource.Datasource) bool {
+	otherTD, ok := other.(*TDengine)
+	if !ok {
+		return false
+	}
+
+	if td.Addr != otherTD.Addr {
+		return false
+	}
+
+	if td.Basic != nil && otherTD.Basic != nil {
+		if td.Basic.User != otherTD.Basic.User {
+			return false
+		}
+
+		if td.Basic.Password != otherTD.Basic.Password {
+			return false
+		}
+	}
+
+	if td.Token != otherTD.Token {
+		return false
+	}
+
+	if td.Timeout != otherTD.Timeout {
+		return false
+	}
+
+	if td.DialTimeout != otherTD.DialTimeout {
+		return false
+	}
+
+	if td.MaxIdleConnsPerHost != otherTD.MaxIdleConnsPerHost {
+		return false
+	}
+
+	if len(td.Headers) != len(otherTD.Headers) {
+		return false
+	}
+
+	for k, v := range td.Headers {
+		if otherV, ok := otherTD.Headers[k]; !ok || v != otherV {
+			return false
+		}
+	}
+	return true
+}
+
+func (td *TDengine) Validate(ctx context.Context) (err error) {
+	return nil
+}
+
+func (td *TDengine) MakeLogQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	return nil, nil
+}
+
+func (td *TDengine) MakeTSQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
+	return nil, nil
+}
+
+func (td *TDengine) QueryData(ctx context.Context, queryParam interface{}) ([]models.DataResp, error) {
+	return td.Query(queryParam, 0)
+}
+
+func (td *TDengine) QueryLog(ctx context.Context, queryParam interface{}) ([]interface{}, int64, error) {
+	b, err := json.Marshal(queryParam)
+	if err != nil {
+		return nil, 0, err
+	}
+	var q TdengineQuery
+	err = json.Unmarshal(b, &q)
+	if err != nil {
+		return nil, 0, err
+	}
+
+	if q.Interval == 0 {
+		q.Interval = 60
+	}
+
+	if q.From == "" {
+		// 2023-09-21T05:37:30.000Z format
+		to := time.Now().Unix()
+		q.To = time.Unix(to, 0).UTC().Format(time.RFC3339)
+		from := to - q.Interval
+		q.From = time.Unix(from, 0).UTC().Format(time.RFC3339)
+	}
+
+	replacements := map[string]string{
+		"$from":     fmt.Sprintf("'%s'", q.From),
+		"$to":       fmt.Sprintf("'%s'", q.To),
+		"$interval": fmt.Sprintf("%ds", q.Interval),
+	}
+
+	for key, val := range replacements {
+		q.Query = strings.ReplaceAll(q.Query, key, val)
+	}
+
+	if !strings.Contains(q.Query, "limit") {
+		q.Query = q.Query + " limit 200"
+	}
+
+	data, err := td.QueryTable(q.Query)
+	if err != nil {
+		return nil, 0, err
+	}
+
+	return ConvertToTable(data), int64(len(data.Data)), nil
+}
+
+func (td *TDengine) QueryMapData(ctx context.Context, query interface{}) ([]map[string]string, error) {
+	return nil, nil
+}
+
+func (td *TDengine) Query(query interface{}, delay ...int) ([]models.DataResp, error) {
+	b, err := json.Marshal(query)
+	if err != nil {
+		return nil, err
+	}
+	var q TdengineQuery
+	err = json.Unmarshal(b, &q)
+	if err != nil {
+		return nil, err
+	}
+
+	if q.Interval == 0 {
+		q.Interval = 60
+	}
+
+	delaySec := 0
+	if len(delay) > 0 {
+		delaySec = delay[0]
+	}
+
+	if q.From == "" {
+		// 2023-09-21T05:37:30.000Z format
+		to := time.Now().Unix() - int64(delaySec)
+		q.To = time.Unix(to, 0).UTC().Format(time.RFC3339)
+		from := to - q.Interval
+		q.From = time.Unix(from, 0).UTC().Format(time.RFC3339)
+	}
+
+	replacements := map[string]string{
+		"$from":     fmt.Sprintf("'%s'", q.From),
+		"$to":       fmt.Sprintf("'%s'", q.To),
+		"$interval": fmt.Sprintf("%ds", q.Interval),
+	}
+
+	for key, val := range replacements {
+		q.Query = strings.ReplaceAll(q.Query, key, val)
+	}
+
+	data, err := td.QueryTable(q.Query)
+	if err != nil {
+		return nil, err
+	}
+	logger.Debugf("tdengine query:%s result: %+v", q.Query, data)
+
+	return ConvertToTStData(data, q.Keys, q.Ref)
+}
+
+func ConvertToTStData(src td.APIResponse, key Keys, ref string) ([]models.DataResp, error) {
+	metricIdxMap := make(map[string]int)
+	labelIdxMap := make(map[string]int)
+
+	metricMap := make(map[string]struct{})
+	if key.MetricKey != "" {
+		metricList := strings.Split(key.MetricKey, " ")
+		for _, metric := range metricList {
+			metricMap[metric] = struct{}{}
+		}
+	}
+
+	labelMap := make(map[string]string)
+	if key.LabelKey != "" {
+		labelList := strings.Split(key.LabelKey, " ")
+		for _, label := range labelList {
+			labelMap[label] = label
+		}
+	}
+
+	tsIdx := -1
+	for colIndex, colData := range src.ColumnMeta {
+		colName := colData[0].(string)
+		var colType string
+		// 处理v2版本数字类型和v3版本字符串类型
+		switch t := colData[1].(type) {
+		case float64:
+			// v2版本数字类型映射
+			switch int(t) {
+			case 1:
+				colType = "BOOL"
+			case 2:
+				colType = "TINYINT"
+			case 3:
+				colType = "SMALLINT"
+			case 4:
+				colType = "INT"
+			case 5:
+				colType = "BIGINT"
+			case 6:
+				colType = "FLOAT"
+			case 7:
+				colType = "DOUBLE"
+			case 8:
+				colType = "BINARY"
+			case 9:
+				colType = "TIMESTAMP"
+			case 10:
+				colType = "NCHAR"
+			default:
+				colType = "UNKNOWN"
+			}
+		case string:
+			// v3版本直接使用字符串类型
+			colType = t
+		default:
+			logger.Warningf("unexpected column type format: %v", colData[1])
+			continue
+		}
+
+		switch colType {
+		case "TIMESTAMP":
+			tsIdx = colIndex
+		case "BIGINT", "INT", "INT UNSIGNED", "BIGINT UNSIGNED", "FLOAT", "DOUBLE",
+			"SMALLINT", "SMALLINT UNSIGNED", "TINYINT", "TINYINT UNSIGNED", "BOOL":
+			if len(metricMap) > 0 {
+				if _, ok := metricMap[colName]; !ok {
+					continue
+				}
+				metricIdxMap[colName] = colIndex
+			} else {
+				metricIdxMap[colName] = colIndex
+			}
+		default:
+			if len(labelMap) > 0 {
+				if _, ok := labelMap[colName]; !ok {
+					continue
+				}
+				labelIdxMap[colName] = colIndex
+			} else {
+				labelIdxMap[colName] = colIndex
+			}
+		}
+	}
+
+	if tsIdx == -1 {
+		return nil, fmt.Errorf("timestamp column not found, please check your query")
+	}
+
+	var result []models.DataResp
+	m := make(map[string]*models.DataResp)
+	for _, row := range src.Data {
+		for metricName, metricIdx := range metricIdxMap {
+			value, err := interfaceToFloat64(row[metricIdx])
+			if err != nil {
+				logger.Warningf("parse %v value failed: %v", row, err)
+				continue
+			}
+
+			metric := make(model.Metric)
+			for labelName, labelIdx := range labelIdxMap {
+				metric[model.LabelName(labelName)] = model.LabelValue(row[labelIdx].(string))
+			}
+
+			metric[model.MetricNameLabel] = model.LabelValue(metricName)
+
+			// transfer 2022-06-29T05:52:16.603Z to unix timestamp
+			t, err := parseTimeString(row[tsIdx].(string))
+			if err != nil {
+				logger.Warningf("parse %v timestamp failed: %v", row, err)
+				continue
+			}
+
+			timestamp := t.Unix()
+			if _, ok := m[metric.String()]; !ok {
+				m[metric.String()] = &models.DataResp{
+					Metric: metric,
+					Values: [][]float64{
+						{float64(timestamp), value},
+					},
+				}
+			} else {
+				m[metric.String()].Values = append(m[metric.String()].Values, []float64{float64(timestamp), value})
+			}
+		}
+	}
+
+	for _, v := range m {
+		v.Ref = ref
+		result = append(result, *v)
+	}
+
+	return result, nil
+}
+
+func interfaceToFloat64(input interface{}) (float64, error) {
+	// Check for the kind of the value first
+	if input == nil {
+		return 0, fmt.Errorf("unsupported type: %T", input)
+	}
+
+	kind := reflect.TypeOf(input).Kind()
+	switch kind {
+	case reflect.Float64:
+		return input.(float64), nil
+	case reflect.Float32:
+		return float64(input.(float32)), nil
+	case reflect.Int, reflect.Int32, reflect.Int64, reflect.Int8, reflect.Int16:
+		return float64(reflect.ValueOf(input).Int()), nil
+	case reflect.Uint, reflect.Uint32, reflect.Uint64, reflect.Uint8, reflect.Uint16:
+		return float64(reflect.ValueOf(input).Uint()), nil
+	case reflect.String:
+		return strconv.ParseFloat(input.(string), 64)
+	case reflect.Bool:
+		if input.(bool) {
+			return 1.0, nil
+		}
+		return 0.0, nil
+	default:
+		return 0, fmt.Errorf("unsupported type: %T", input)
+	}
+}
+
+func parseTimeString(ts string) (time.Time, error) {
+	// 尝试不同的时间格式
+	formats := []string{
+		// 标准格式
+		time.Layout,      // "01/02 03:04:05PM '06 -0700"
+		time.ANSIC,       // "Mon Jan _2 15:04:05 2006"
+		time.UnixDate,    // "Mon Jan _2 15:04:05 MST 2006"
+		time.RubyDate,    // "Mon Jan 02 15:04:05 -0700 2006"
+		time.RFC822,      // "02 Jan 06 15:04 MST"
+		time.RFC822Z,     // "02 Jan 06 15:04 -0700"
+		time.RFC850,      // "Monday, 02-Jan-06 15:04:05 MST"
+		time.RFC1123,     // "Mon, 02 Jan 2006 15:04:05 MST"
+		time.RFC1123Z,    // "Mon, 02 Jan 2006 15:04:05 -0700"
+		time.RFC3339,     // "2006-01-02T15:04:05Z07:00"
+		time.RFC3339Nano, // "2006-01-02T15:04:05.999999999Z07:00"
+		time.Kitchen,     // "3:04PM"
+
+		// 实用时间戳格式
+		time.Stamp,      // "Jan _2 15:04:05"
+		time.StampMilli, // "Jan _2 15:04:05.000"
+		time.StampMicro, // "Jan _2 15:04:05.000000"
+		time.StampNano,  // "Jan _2 15:04:05.000000000"
+		time.DateTime,   // "2006-01-02 15:04:05"
+		time.DateOnly,   // "2006-01-02"
+		time.TimeOnly,   // "15:04:05"
+
+		// 常用自定义格式
+		"2006-01-02T15:04:05", // 无时区的ISO格式
+		"2006-01-02T15:04:05.000Z",
+		"2006-01-02T15:04:05Z",
+		"2006-01-02 15:04:05.999999999", // 纳秒
+		"2006-01-02 15:04:05.999999",    // 微秒
+		"2006-01-02 15:04:05.999",       // 毫秒
+		"2006/01/02",
+		"20060102",
+		"01/02/2006",
+		"2006年01月02日",
+		"2006年01月02日 15:04:05",
+	}
+
+	var lastErr error
+	for _, format := range formats {
+		t, err := time.Parse(format, ts)
+		if err == nil {
+			return t, nil
+		}
+		lastErr = err
+	}
+
+	// 尝试解析 Unix 时间戳
+	if timestamp, err := strconv.ParseInt(ts, 10, 64); err == nil {
+		switch len(ts) {
+		case 10: // 秒
+			return time.Unix(timestamp, 0), nil
+		case 13: // 毫秒
+			return time.Unix(timestamp/1000, (timestamp%1000)*1000000), nil
+		case 16: // 微秒
+			return time.Unix(timestamp/1000000, (timestamp%1000000)*1000), nil
+		case 19: // 纳秒
+			return time.Unix(timestamp/1000000000, timestamp%1000000000), nil
+		}
+	}
+
+	return time.Time{}, fmt.Errorf("failed to parse time with any format: %v", lastErr)
+}
+
+func ConvertToTable(src td.APIResponse) []interface{} {
+	var resp []interface{}
+
+	for i := range src.Data {
+		cur := make(map[string]interface{})
+		for j := range src.Data[i] {
+			cur[src.ColumnMeta[j][0].(string)] = src.Data[i][j]
+		}
+		resp = append(resp, cur)
+	}
+	return resp
+}

docker/compose-bridge/docker-compose.yaml

@@ -1,5 +1,3 @@
-version: "3.7"
-
 networks:
   nightingale:
     driver: bridge

docker/compose-bridge/etc-categraf/config.toml

@@ -19,8 +19,8 @@ precision = "ms"
 # global collect interval
 interval = 15
 
-[global.labels]
-source="categraf"
+# [global.labels]
+# source="categraf"
 # region = "shanghai"
 # env = "localhost"
 

docker/compose-bridge/etc-categraf/input.mysql/mysql.toml

@@ -0,0 +1,42 @@
+[[instances]]
+address = "mysql:3306"
+username = "root"
+password = "1234"
+
+# # set tls=custom to enable tls
+# parameters = "tls=false"
+
+# extra_status_metrics = true
+# extra_innodb_metrics = false
+# gather_processlist_processes_by_state = false
+# gather_processlist_processes_by_user = false
+# gather_schema_size = true
+# gather_table_size = false
+# gather_system_table_size = false
+# gather_slave_status = true
+
+# # timeout
+# timeout_seconds = 3
+
+# # interval = global.interval * interval_times
+# interval_times = 1
+
+# important! use global unique string to specify instance
+labels = { instance="docker-compose-mysql" }
+
+## Optional TLS Config
+# use_tls = false
+# tls_min_version = "1.2"
+# tls_ca = "/etc/categraf/ca.pem"
+# tls_cert = "/etc/categraf/cert.pem"
+# tls_key = "/etc/categraf/key.pem"
+## Use TLS but skip chain & host verification
+# insecure_skip_verify = true
+
+#[[instances.queries]]
+# mesurement = "lock_wait"
+# metric_fields = [ "total" ]
+# timeout = "3s"
+# request = '''
+#SELECT count(*) as total FROM information_schema.innodb_trx WHERE trx_state='LOCK WAIT'
+#'''

docker/compose-bridge/etc-categraf/input.redis/redis.toml

@@ -0,0 +1,37 @@
+[[instances]]
+address = "redis:6379"
+username = ""
+password = ""
+# pool_size = 2
+
+## 是否开启slowlog 收集
+# gather_slowlog = true
+## 最多收集少条slowlog
+# slowlog_max_len = 100
+## 收集距离现在多少秒以内的slowlog
+## 注意插件的采集周期,该参数不要小于采集周期，否则会有slowlog查不到
+# slowlog_time_window=30
+
+# 指标
+# redis_slow_log{ident=dev-01 client_addr=127.0.0.1:56364 client_name= cmd="info ALL" log_id=983} 74 (单位微秒)
+
+# # Optional. Specify redis commands to retrieve values
+# commands = [
+#     {command = ["get", "sample-key1"], metric = "custom_metric_name1"},
+#     {command = ["get", "sample-key2"], metric = "custom_metric_name2"}
+# ]
+
+# # interval = global.interval * interval_times
+# interval_times = 1
+
+# important! use global unique string to specify instance
+labels = { instance="docker-compose-redis" }
+
+## Optional TLS Config
+# use_tls = false
+# tls_min_version = "1.2"
+# tls_ca = "/etc/categraf/ca.pem"
+# tls_cert = "/etc/categraf/cert.pem"
+# tls_key = "/etc/categraf/key.pem"
+## Use TLS but skip chain & host verification
+# insecure_skip_verify = true

docker/compose-bridge/etc-nightingale/config.toml

@@ -50,7 +50,7 @@ Enable = true
 # user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 
 [HTTP.APIForService]
-Enable = true 
+Enable = false
 [HTTP.APIForService.BasicAuth]
 user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 

docker/compose-host-network-metric-log/etc-nightingale/config.toml

@@ -50,7 +50,7 @@ Enable = true
 # user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 
 [HTTP.APIForService]
-Enable = true 
+Enable = false
 [HTTP.APIForService.BasicAuth]
 user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 

docker/compose-host-network/docker-compose.yaml

@@ -25,7 +25,7 @@ services:
     network_mode: host
 
   prometheus:
-    image: prom/prometheus
+    image: prom/prometheus:v2.55.1
     container_name: prometheus
     hostname: prometheus
     restart: always

docker/compose-host-network/etc-nightingale/config.toml

@@ -50,7 +50,7 @@ Enable = true
 # user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 
 [HTTP.APIForService]
-Enable = true 
+Enable = false
 [HTTP.APIForService.BasicAuth]
 user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 

docker/compose-postgres/initsql_for_postgres/a-n9e-for-Postgres.sql

@@ -561,7 +561,7 @@ CREATE TABLE alert_cur_event (
     target_note varchar(191) not null default '' ,
     first_trigger_time bigint,
     trigger_time bigint not null,
-    trigger_value varchar(255) not null,
+    trigger_value varchar(2048) not null,
     annotations text not null ,
     rule_config text not null ,
     tags varchar(1024) not null default '' ,
@@ -621,7 +621,7 @@ CREATE TABLE alert_his_event (
     target_note varchar(191) not null default '' ,
     first_trigger_time bigint,
     trigger_time bigint not null,
-    trigger_value varchar(255) not null,
+    trigger_value varchar(2048) not null,
     recover_time bigint not null default 0,
     last_eval_time bigint not null default 0 ,
     tags varchar(1024) not null default '' ,
@@ -790,6 +790,7 @@ CREATE TABLE es_index_pattern (
     time_field varchar(128) not null default '@timestamp',
     allow_hide_system_indices smallint not null default 0,
     fields_format varchar(4096) not null default '',
+    cross_cluster_enabled int not null default 0,
     create_at bigint default '0',
     create_by varchar(64) default '',
     update_at bigint default '0',
@@ -859,6 +860,7 @@ CREATE TABLE builtin_components (
   ident VARCHAR(191) NOT NULL,
   logo VARCHAR(191) NOT NULL,
   readme TEXT NOT NULL,
+  disabled INT NOT NULL DEFAULT 0,
   created_at BIGINT NOT NULL DEFAULT 0,
   created_by VARCHAR(191) NOT NULL DEFAULT '',
   updated_at BIGINT NOT NULL DEFAULT 0,
@@ -885,4 +887,20 @@ CREATE TABLE builtin_payloads (
 CREATE INDEX idx_component ON builtin_payloads (component);
 CREATE INDEX idx_builtin_payloads_name ON builtin_payloads (name);
 CREATE INDEX idx_cate ON builtin_payloads (cate);
-CREATE INDEX idx_type ON builtin_payloads (type);
+CREATE INDEX idx_type ON builtin_payloads (type);
+
+
+CREATE TABLE dash_annotation (
+    id bigserial PRIMARY KEY,
+    dashboard_id bigint not null,
+    panel_id varchar(191) not null,
+    tags text,
+    description text,
+    config text,
+    time_start bigint not null default 0,
+    time_end bigint not null default 0,
+    create_at bigint not null default 0,
+    create_by varchar(64) not null default '',
+    update_at bigint not null default 0,
+    update_by varchar(64) not null default ''
+);

docker/compose-postgres/n9eetc_pg/config.toml

@@ -50,7 +50,7 @@ Enable = true
 # user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 
 [HTTP.APIForService]
-Enable = true 
+Enable = false
 [HTTP.APIForService.BasicAuth]
 user001 = "ccc26da7b9aba533cbb263a36c07dcc5"
 

docker/initsql/c-init.sql

@@ -1,3 +1,10 @@
-GRANT ALL ON *.* TO 'root'@'127.0.0.1' IDENTIFIED BY '1234';
-GRANT ALL ON *.* TO 'root'@'localhost' IDENTIFIED BY '1234';
-GRANT ALL ON *.* TO 'root'@'%' IDENTIFIED BY '1234';
+CREATE USER IF NOT EXISTS 'root'@'127.0.0.1' IDENTIFIED BY '1234';
+GRANT ALL PRIVILEGES ON *.* TO 'root'@'127.0.0.1' WITH GRANT OPTION;
+
+CREATE USER IF NOT EXISTS 'root'@'localhost' IDENTIFIED BY '1234';
+GRANT ALL PRIVILEGES ON *.* TO 'root'@'localhost' WITH GRANT OPTION;
+
+CREATE USER IF NOT EXISTS 'root'@'%' IDENTIFIED BY '1234';
+GRANT ALL PRIVILEGES ON *.* TO 'root'@'%' WITH GRANT OPTION;
+
+FLUSH PRIVILEGES;

docker/migratesql/migrate.sql

@@ -83,4 +83,68 @@ ALTER TABLE recording_rule ADD COLUMN cron_pattern VARCHAR(255) DEFAULT '' COMME
 
 /* v7.0.0-beta.14 */
 ALTER TABLE alert_cur_event ADD COLUMN original_tags TEXT COMMENT 'labels key=val,,k2=v2';
-ALTER TABLE alert_his_event ADD COLUMN original_tags TEXT COMMENT 'labels key=val,,k2=v2';
+ALTER TABLE alert_his_event ADD COLUMN original_tags TEXT COMMENT 'labels key=val,,k2=v2';
+
+/* v7.1.0 */
+ALTER TABLE target ADD COLUMN os VARCHAR(31) DEFAULT '' COMMENT 'os type';
+
+/* v7.2.0 */
+CREATE TABLE notification_record (
+    `id` BIGINT PRIMARY KEY AUTO_INCREMENT,
+    `event_id` BIGINT NOT NULL,
+    `sub_id` BIGINT NOT NULL,
+    `channel` VARCHAR(255) NOT NULL,
+    `status` TINYINT NOT NULL DEFAULT 0,
+    `target` VARCHAR(1024) NOT NULL,
+    `details` VARCHAR(2048),
+    `created_at` BIGINT NOT NULL,
+    INDEX idx_evt (event_id)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;
+
+
+/* v7.3.0 2024-08-26 */
+ALTER TABLE `target` ADD COLUMN `host_tags` TEXT COMMENT 'global labels set in conf file';
+
+/* v7.3.4 2024-08-28 */
+ALTER TABLE `builtin_payloads` ADD COLUMN `component_id` bigint(20) NOT NULL DEFAULT 0 COMMENT 'component_id';
+
+/* v7.4.0 2024-09-20 */
+CREATE TABLE `target_busi_group` (
+  `id` bigint NOT NULL AUTO_INCREMENT,
+  `target_ident` varchar(191) NOT NULL,
+  `group_id` bigint NOT NULL,
+  `update_at` bigint NOT NULL,
+  PRIMARY KEY (`id`),
+  UNIQUE KEY `idx_target_group` (`target_ident`,`group_id`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;
+
+/* v7.7.2 2024-12-02 */
+ALTER TABLE alert_subscribe MODIFY COLUMN rule_ids varchar(1024);
+ALTER TABLE alert_subscribe MODIFY COLUMN busi_groups varchar(4096);
+
+/* v8.0.0-beta.1 2024-12-13 */
+ALTER TABLE `alert_rule` ADD COLUMN `cron_pattern` VARCHAR(64);
+ALTER TABLE `builtin_components` MODIFY COLUMN `logo` mediumtext COMMENT '''logo of component''';
+
+/* v8.0.0-beta.2 2024-12-26 */
+ALTER TABLE `es_index_pattern` ADD COLUMN `cross_cluster_enabled` int not null default 0;
+
+/* v8.0.0-beta.3 2024-01-03 */
+ALTER TABLE `builtin_components` ADD COLUMN `disabled` INT NOT NULL DEFAULT 0 COMMENT 'is disabled or not';
+        
+CREATE TABLE `dash_annotation` (
+    `id` bigint unsigned not null auto_increment,
+    `dashboard_id` bigint not null comment 'dashboard id',
+    `panel_id` varchar(191) not null comment 'panel id',
+    `tags` text comment 'tags array json string',
+    `description` text comment 'annotation description',
+    `config` text comment 'annotation config',
+    `time_start` bigint not null default 0 comment 'start timestamp',
+    `time_end` bigint not null default 0 comment 'end timestamp',
+    `create_at` bigint not null default 0 comment 'create time',
+    `create_by` varchar(64) not null default '' comment 'creator',
+    `update_at` bigint not null default 0 comment 'update time',
+    `update_by` varchar(64) not null default '' comment 'updater',
+    PRIMARY KEY (`id`),
+    KEY `idx_dashboard_id` (`dashboard_id`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;

docker/sqlite.sql

@@ -17,6 +17,8 @@ CREATE TABLE `users` (
     `update_by` varchar(64) not null default ''
 );
 
+CREATE UNIQUE INDEX idx_users_username ON `users` (username);
+
 insert into `users`(id, username, nickname, password, roles, create_at, create_by, update_at, update_by) values(1, 'root', '超管', 'root.2020', 'Admin', strftime('%s', 'now'), 'system', strftime('%s', 'now'), 'system');
 
 CREATE TABLE `user_group` (
@@ -182,8 +184,9 @@ CREATE TABLE `board` (
     `create_by` varchar(64) not null default '',
     `update_at` bigint not null default 0,
     `update_by` varchar(64) not null default '',
-    unique (`group_id`, `name`)
+    `public_cate` bigint not null default 0
 );
+CREATE UNIQUE INDEX idx_board_group_id_name ON `board` (group_id, name);
 CREATE INDEX `idx_board_ident` ON `board` (`ident` asc);
 
 -- for dashboard new version
@@ -192,6 +195,15 @@ CREATE TABLE `board_payload` (
     `payload` mediumtext not null
 );
 
+CREATE TABLE `chart` (
+    `id` integer primary key autoincrement,
+    `group_id` integer not null,
+    `configs` text,
+    `weight` integer not null default 0
+);
+
+CREATE INDEX idx_chart_group_id ON `chart` (group_id);
+
 CREATE TABLE `chart_share` (
     `id` integer primary key autoincrement,
     `cluster` varchar(128) not null,
@@ -238,7 +250,9 @@ CREATE TABLE `alert_rule` (
     `create_at` bigint not null default 0,
     `create_by` varchar(64) not null default '',
     `update_at` bigint not null default 0,
-    `update_by` varchar(64) not null default ''
+    `update_by` varchar(64) not null default '',
+    `cron_pattern` varchar(64),
+    `datasource_queries` text
 );
 CREATE INDEX `idx_alert_rule_group_id` ON `alert_rule` (`group_id` asc);
 CREATE INDEX `idx_alert_rule_update_at` ON `alert_rule` (`update_at` asc);
@@ -308,11 +322,18 @@ CREATE TABLE `target` (
     `tags` varchar(512) not null default '',
     `host_ip` varchar(15) default '',
     `agent_version` varchar(255) default '',
+    `host_tags` text,
     `engine_name` varchar(255) default '',
+    `os` varchar(31) default '',
     `update_at` bigint not null default 0
 );
-CREATE INDEX `idx_target_group_id` ON `target` (`group_id` asc);
 
+CREATE INDEX `idx_target_group_id` ON `target` (`group_id` asc);
+CREATE UNIQUE INDEX idx_target_ident ON `target` (ident);
+CREATE INDEX idx_host_ip ON `target` (host_ip);
+CREATE INDEX idx_agent_version ON `target` (agent_version);
+CREATE INDEX idx_engine_name ON `target` (engine_name);
+CREATE INDEX idx_os ON `target` (os);
 
 CREATE TABLE `metric_view` (
     `id` integer primary key autoincrement,
@@ -337,12 +358,14 @@ CREATE TABLE `recording_rule` (
     `disabled` tinyint(1) not null default 0,
     `prom_ql` varchar(8192) not null,
     `prom_eval_interval` int not null,
+    `cron_pattern` varchar(255) default '',
     `append_tags` varchar(255) default '',
     `query_configs` text not null,
     `create_at` bigint default '0',
     `create_by` varchar(64) default '',
     `update_at` bigint default '0',
-    `update_by` varchar(64) default ''
+    `update_by` varchar(64) default '',
+    `datasource_queries` text
 );
 CREATE INDEX `idx_recording_rule_group_id` ON `recording_rule` (`group_id` asc);
 CREATE INDEX `idx_recording_rule_update_at` ON `recording_rule` (`update_at` asc);
@@ -389,7 +412,7 @@ CREATE TABLE `alert_cur_event` (
     `target_note` varchar(191) not null default '',
     `first_trigger_time` bigint,
     `trigger_time` bigint not null,
-    `trigger_value` varchar(255) not null,
+    `trigger_value` varchar(2048) not null,
     `annotations` text not null,
     `rule_config` text not null,
     `tags` varchar(1024) not null default ''
@@ -427,9 +450,10 @@ CREATE TABLE `alert_his_event` (
     `target_note` varchar(191) not null default '',
     `first_trigger_time` bigint,
     `trigger_time` bigint not null,
-    `trigger_value` varchar(255) not null,
+    `trigger_value` varchar(2048) not null,
     `recover_time` bigint not null default 0,
     `last_eval_time` bigint not null default 0,
+    `original_tags` varchar(8192),
     `tags` varchar(1024) not null default '',
     `annotations` text not null,
     `rule_config` text not null
@@ -459,6 +483,8 @@ CREATE INDEX `idx_builtin_components_ident` ON `builtin_components` (`ident` asc
 
 CREATE TABLE `builtin_payloads` (
   `id` integer primary key autoincrement,
+  `component_id` integer not null default 0,
+  `uuid` integer not null,
   `type` varchar(191) not null,
   `component` varchar(191) not null,
   `cate` varchar(191) not null,
@@ -474,6 +500,20 @@ CREATE INDEX `idx_builtin_payloads_component` ON `builtin_payloads` (`component`
 CREATE INDEX `idx_builtin_payloads_name` ON `builtin_payloads` (`name` asc);
 CREATE INDEX `idx_builtin_payloads_cate` ON `builtin_payloads` (`cate` asc);
 CREATE INDEX `idx_builtin_payloads_type` ON `builtin_payloads` (`type` asc);
+CREATE INDEX idx_uuid ON `builtin_payloads` (uuid);
+
+
+CREATE TABLE `notification_record` (
+    `id` integer primary key autoincrement,
+    `event_id` integer not null,
+    `sub_id` integer,
+    `channel` varchar(255) not null,
+    `status` integer,
+    `target` varchar(1024) not null,
+    `details` varchar(2048) default '',
+    `created_at` integer not null
+);
+CREATE INDEX idx_evt ON notification_record (event_id);
 
 CREATE TABLE `task_tpl` (
     `id`        integer primary key autoincrement,
@@ -553,6 +593,8 @@ CREATE TABLE `datasource`
     `updated_by` varchar(64) not null default ''
 );
 
+CREATE UNIQUE INDEX idx_datasource_name ON datasource (name);
+
 CREATE TABLE `builtin_cate` (
     `id` integer primary key autoincrement,
     `name` varchar(191) not null,
@@ -570,6 +612,8 @@ CREATE TABLE `notify_tpl` (
     `update_by` varchar(64) not null default ''
 );
 
+CREATE UNIQUE INDEX idx_notify_tpl_channel ON notify_tpl (channel);
+
 CREATE TABLE `sso_config` (
     `id` integer primary key autoincrement,
     `name` varchar(191) not null unique,
@@ -577,6 +621,8 @@ CREATE TABLE `sso_config` (
     `update_at` bigint not null default 0
 );
 
+CREATE UNIQUE INDEX idx_sso_config_name ON sso_config (name);
+
 CREATE TABLE `es_index_pattern` (
     `id` integer primary key autoincrement,
     `datasource_id` bigint not null default 0,
@@ -584,6 +630,7 @@ CREATE TABLE `es_index_pattern` (
     `time_field` varchar(128) not null default '@timestamp',
     `allow_hide_system_indices` tinyint(1) not null default 0,
     `fields_format` varchar(4096) not null default '',
+    `cross_cluster_enabled` int not null default 0,
     `create_at` bigint default '0',
     `create_by` varchar(64) default '',
     `update_at` bigint default '0',
@@ -591,6 +638,8 @@ CREATE TABLE `es_index_pattern` (
     unique (`datasource_id`, `name`)
 );
 
+CREATE UNIQUE INDEX idx_es_index_pattern_datasource_id_name ON es_index_pattern (datasource_id, name);
+
 CREATE TABLE `builtin_metrics` (
     `id` integer primary key autoincrement,
     `collector` varchar(191) NOT NULL,
@@ -603,13 +652,15 @@ CREATE TABLE `builtin_metrics` (
     `created_at` bigint NOT NULL DEFAULT 0,
     `created_by` varchar(191) NOT NULL DEFAULT '',
     `updated_at` bigint NOT NULL DEFAULT 0,
-    `updated_by` varchar(191) NOT NULL DEFAULT ''
+    `updated_by` varchar(191) NOT NULL DEFAULT '',
+    `uuid integer` not null default 0
 );
--- CREATE UNIQUE INDEX `idx_builtin_metrics_collector_typ_name` ON `builtin_metrics` (`lang`,`collector`, `typ`, `name` asc);
--- CREATE INDEX `idx_builtin_metrics_collector` ON `builtin_metrics` (`collector` asc);
--- CREATE INDEX `idx_builtin_metrics_typ` ON `builtin_metrics` (`typ` asc);
--- CREATE INDEX `idx_builtin_metrics_name` ON `builtin_metrics` (`name` asc);
--- CREATE INDEX `idx_builtin_metrics_lang` ON `builtin_metrics` (`lang` asc);
+
+CREATE UNIQUE INDEX idx_collector_typ_name ON builtin_metrics (lang, collector, typ, name);
+CREATE INDEX idx_collector ON builtin_metrics (collector);
+CREATE INDEX idx_typ ON builtin_metrics (typ);
+CREATE INDEX idx_builtinmetric_name ON builtin_metrics (name);
+CREATE INDEX idx_lang ON builtin_metrics (lang);
 
 
 CREATE TABLE `metric_filter` (
@@ -624,6 +675,30 @@ CREATE TABLE `metric_filter` (
 );
 CREATE INDEX `idx_metric_filter_name` ON `metric_filter` (`name` asc);
 
+CREATE TABLE `target_busi_group` (
+    `id` integer primary key autoincrement,
+    `target_ident` varchar(191) not null,
+    `group_id` integer not null,
+    `update_at` integer not null
+);
+
+CREATE UNIQUE INDEX idx_target_busi_group ON target_busi_group (target_ident, group_id);
+
+
+CREATE TABLE `dash_annotation` (
+    `id` integer primary key autoincrement,
+    `dashboard_id` bigint not null,
+    `panel_id` varchar(191) not null,
+    `tags` text,
+    `description` text,
+    `config` text,
+    `time_start` bigint not null default 0,
+    `time_end` bigint not null default 0,
+    `create_at` bigint not null default 0,
+    `create_by` varchar(64) not null default '',
+    `update_at` bigint not null default 0,
+    `update_by` varchar(64) not null default ''
+);
 
 CREATE TABLE `task_meta`
 (

dscache/cache.go

@@ -0,0 +1,59 @@
+package dscache
+
+import (
+	"sync"
+
+	"github.com/ccfos/nightingale/v6/datasource"
+	"github.com/toolkits/pkg/logger"
+)
+
+type Cache struct {
+	datas map[string]map[int64]datasource.Datasource
+	mutex *sync.RWMutex
+}
+
+var DsCache = Cache{
+	datas: make(map[string]map[int64]datasource.Datasource),
+	mutex: new(sync.RWMutex),
+}
+
+func (cs *Cache) Put(cate string, dsId int64, ds datasource.Datasource) {
+	cs.mutex.Lock()
+	if _, found := cs.datas[cate]; !found {
+		cs.datas[cate] = make(map[int64]datasource.Datasource)
+	}
+
+	if _, found := cs.datas[cate][dsId]; found {
+		if cs.datas[cate][dsId].Equal(ds) {
+			cs.mutex.Unlock()
+			return
+		}
+	}
+	cs.mutex.Unlock()
+
+	// InitClient() 在用户配置错误或远端不可用时, 会非常耗时, mutex被长期持有, 导致Get()会超时
+	err := ds.InitClient()
+	if err != nil {
+		logger.Errorf("init plugin:%s %d %+v client fail: %v", cate, dsId, ds, err)
+		return
+	}
+
+	logger.Debugf("init plugin:%s %d %+v client success", cate, dsId, ds)
+	cs.mutex.Lock()
+	cs.datas[cate][dsId] = ds
+	cs.mutex.Unlock()
+}
+
+func (cs *Cache) Get(cate string, dsId int64) (datasource.Datasource, bool) {
+	cs.mutex.RLock()
+	defer cs.mutex.RUnlock()
+	if _, found := cs.datas[cate]; !found {
+		return nil, false
+	}
+
+	if _, found := cs.datas[cate][dsId]; !found {
+		return nil, false
+	}
+
+	return cs.datas[cate][dsId], true
+}

dscache/sync.go

@@ -0,0 +1,190 @@
+package dscache
+
+import (
+	"context"
+	"strings"
+	"sync/atomic"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/datasource"
+	_ "github.com/ccfos/nightingale/v6/datasource/ck"
+	"github.com/ccfos/nightingale/v6/datasource/es"
+	"github.com/ccfos/nightingale/v6/dskit/tdengine"
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/ctx"
+
+	"github.com/toolkits/pkg/logger"
+)
+
+var FromAPIHook func()
+
+func Init(ctx *ctx.Context, fromAPI bool) {
+	go getDatasourcesFromDBLoop(ctx, fromAPI)
+}
+
+type ListInput struct {
+	Page       int    `json:"p"`
+	Limit      int    `json:"limit"`
+	Category   string `json:"category"`
+	PluginType string `json:"plugin_type"` // promethues
+	Status     string `json:"status"`
+}
+
+type DSReply struct {
+	RequestID string `json:"request_id"`
+	Data      struct {
+		Items []datasource.DatasourceInfo `json:"items"`
+	} `json:"data"`
+}
+
+type DSReplyEncrypt struct {
+	RequestID string `json:"request_id"`
+	Data      string `json:"data"`
+}
+
+var PromDefaultDatasourceId int64
+
+func getDatasourcesFromDBLoop(ctx *ctx.Context, fromAPI bool) {
+	for {
+		if !fromAPI {
+			items, err := models.GetDatasources(ctx)
+			if err != nil {
+				logger.Errorf("get datasource from database fail: %v", err)
+				//stat.CounterExternalErrorTotal.WithLabelValues("db", "get_cluster").Inc()
+				time.Sleep(time.Second * 2)
+				continue
+			}
+			var dss []datasource.DatasourceInfo
+			for _, item := range items {
+				if item.PluginType == "prometheus" && item.IsDefault {
+					atomic.StoreInt64(&PromDefaultDatasourceId, item.Id)
+				}
+
+				logger.Debugf("get datasource: %+v", item)
+				ds := datasource.DatasourceInfo{
+					Id:             item.Id,
+					Name:           item.Name,
+					Description:    item.Description,
+					Category:       item.Category,
+					PluginId:       item.PluginId,
+					Type:           item.PluginType,
+					PluginTypeName: item.PluginTypeName,
+					Settings:       item.SettingsJson,
+					HTTPJson:       item.HTTPJson,
+					AuthJson:       item.AuthJson,
+					Status:         item.Status,
+					IsDefault:      item.IsDefault,
+				}
+
+				if item.PluginType == "elasticsearch" {
+					esN9eToDatasourceInfo(&ds, item)
+				} else if item.PluginType == "opensearch" {
+					osN9eToDatasourceInfo(&ds, item)
+				} else if item.PluginType == "tdengine" {
+					tdN9eToDatasourceInfo(&ds, item)
+				} else {
+					ds.Settings = make(map[string]interface{})
+					for k, v := range item.SettingsJson {
+						ds.Settings[k] = v
+					}
+				}
+				dss = append(dss, ds)
+			}
+			PutDatasources(dss)
+		} else {
+			FromAPIHook()
+		}
+
+		time.Sleep(time.Second * 2)
+	}
+}
+
+func tdN9eToDatasourceInfo(ds *datasource.DatasourceInfo, item models.Datasource) {
+	ds.Settings = make(map[string]interface{})
+	ds.Settings["tdengine.cluster_name"] = item.Name
+	ds.Settings["tdengine.addr"] = item.HTTPJson.Url
+	ds.Settings["tdengine.timeout"] = item.HTTPJson.Timeout
+	ds.Settings["tdengine.dial_timeout"] = item.HTTPJson.DialTimeout
+	ds.Settings["tdengine.max_idle_conns_per_host"] = item.HTTPJson.MaxIdleConnsPerHost
+	ds.Settings["tdengine.headers"] = item.HTTPJson.Headers
+	ds.Settings["tdengine.basic"] = tdengine.TDengineBasicAuth{
+		User:     item.AuthJson.BasicAuthUser,
+		Password: item.AuthJson.BasicAuthPassword,
+	}
+}
+
+func esN9eToDatasourceInfo(ds *datasource.DatasourceInfo, item models.Datasource) {
+	ds.Settings = make(map[string]interface{})
+	ds.Settings["es.nodes"] = []string{item.HTTPJson.Url}
+	if len(item.HTTPJson.Urls) > 0 {
+		ds.Settings["es.nodes"] = item.HTTPJson.Urls
+	}
+	ds.Settings["es.timeout"] = item.HTTPJson.Timeout
+	ds.Settings["es.basic"] = es.BasicAuth{
+		Username: item.AuthJson.BasicAuthUser,
+		Password: item.AuthJson.BasicAuthPassword,
+	}
+	ds.Settings["es.tls"] = es.TLS{
+		SkipTlsVerify: item.HTTPJson.TLS.SkipTlsVerify,
+	}
+	ds.Settings["es.version"] = item.SettingsJson["version"]
+	ds.Settings["es.headers"] = item.HTTPJson.Headers
+	ds.Settings["es.min_interval"] = item.SettingsJson["min_interval"]
+	ds.Settings["es.max_shard"] = item.SettingsJson["max_shard"]
+	ds.Settings["es.enable_write"] = item.SettingsJson["enable_write"]
+}
+
+// for opensearch
+func osN9eToDatasourceInfo(ds *datasource.DatasourceInfo, item models.Datasource) {
+	ds.Settings = make(map[string]interface{})
+	ds.Settings["os.nodes"] = []string{item.HTTPJson.Url}
+	ds.Settings["os.timeout"] = item.HTTPJson.Timeout
+	ds.Settings["os.basic"] = es.BasicAuth{
+		Username: item.AuthJson.BasicAuthUser,
+		Password: item.AuthJson.BasicAuthPassword,
+	}
+	ds.Settings["os.tls"] = es.TLS{
+		SkipTlsVerify: item.HTTPJson.TLS.SkipTlsVerify,
+	}
+	ds.Settings["os.version"] = item.SettingsJson["version"]
+	ds.Settings["os.headers"] = item.HTTPJson.Headers
+	ds.Settings["os.min_interval"] = item.SettingsJson["min_interval"]
+	ds.Settings["os.max_shard"] = item.SettingsJson["max_shard"]
+}
+
+func PutDatasources(items []datasource.DatasourceInfo) {
+	ids := make([]int64, 0)
+	for _, item := range items {
+		if item.Type == "prometheus" {
+			continue
+		}
+
+		if item.Type == "loki" {
+			continue
+		}
+
+		if item.Name == "" {
+			logger.Warningf("cluster name is empty, ignore %+v", item)
+			continue
+		}
+		typ := strings.ReplaceAll(item.Type, ".logging", "")
+
+		ds, err := datasource.GetDatasourceByType(typ, item.Settings)
+		if err != nil {
+			logger.Debugf("get plugin:%+v fail: %v", item, err)
+			continue
+		}
+
+		err = ds.Validate(context.Background())
+		if err != nil {
+			logger.Warningf("get plugin:%+v fail: %v", item, err)
+			continue
+		}
+		ids = append(ids, item.Id)
+
+		// 异步初始化 client 不然数据源同步的会很慢
+		go DsCache.Put(typ, item.Id, ds)
+	}
+
+	logger.Debugf("get plugin by type success Ids:%v", ids)
+}

dskit/clickhouse/clickhouse.go

@@ -0,0 +1,303 @@
+package clickhouse
+
+import (
+	"context"
+	"database/sql"
+	"errors"
+	"fmt"
+	"io"
+	"strings"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/dskit/sqlbase"
+	"github.com/ccfos/nightingale/v6/dskit/types"
+
+	"github.com/ClickHouse/clickhouse-go/v2"
+	"github.com/mitchellh/mapstructure"
+	"github.com/toolkits/pkg/net/httplib"
+	ckDriver "gorm.io/driver/clickhouse"
+	"gorm.io/gorm"
+)
+
+const (
+	ckDataSource = "clickhouse://%s:%s@%s?read_timeout=10s"
+
+	DefaultLimit = 500
+)
+
+type Clickhouse struct {
+	Nodes        []string `json:"ck.nodes" mapstructure:"ck.nodes"`
+	User         string   `json:"ck.user" mapstructure:"ck.user"`
+	Password     string   `json:"ck.password" mapstructure:"ck.password"`
+	Timeout      int      `json:"ck.timeout" mapstructure:"ck.timeout"`
+	MaxQueryRows int      `json:"ck.max_query_rows" mapstructure:"ck.max_query_rows"`
+
+	Client       *gorm.DB `json:"-"`
+	ClientByHTTP *sql.DB  `json:"-"`
+}
+
+func (c *Clickhouse) InitCli() error {
+	if c.MaxQueryRows == 0 {
+		c.MaxQueryRows = DefaultLimit
+	}
+
+	if len(c.Nodes) == 0 {
+		return fmt.Errorf("not found ck shard, please check datasource config")
+	}
+	addr := c.Nodes[0]
+	url := addr
+	if !strings.HasPrefix(url, "http://") {
+		url = "http://" + url
+	}
+	resp, err := httplib.Get(url).SetTimeout(time.Second * 1).Response()
+	// 忽略HTTP Code错误, 因为可能不是HTTP协议
+	if err != nil {
+		return err
+	}
+	defer resp.Body.Close()
+	// HTTP 协议
+	if resp.StatusCode == 200 {
+		jsonBytes, _ := io.ReadAll(resp.Body)
+		if len(jsonBytes) > 0 && strings.Contains(strings.ToLower(string(jsonBytes)), "ok.") {
+			ckconn := clickhouse.OpenDB(&clickhouse.Options{
+				Addr: []string{addr},
+				Auth: clickhouse.Auth{
+					Username: c.User,
+					Password: c.Password,
+				},
+				Settings: clickhouse.Settings{
+					"max_execution_time": 60,
+				},
+				DialTimeout: 10 * time.Second,
+				Protocol:    clickhouse.HTTP,
+			})
+			if ckconn == nil {
+				return errors.New("db conn failed")
+			}
+			c.ClientByHTTP = ckconn
+			return nil
+		}
+	}
+
+	db, err := gorm.Open(
+		ckDriver.New(
+			ckDriver.Config{
+				DSN: fmt.Sprintf(ckDataSource,
+					c.User, c.Password, addr),
+				DisableDatetimePrecision:  true,
+				DontSupportRenameColumn:   true,
+				SkipInitializeWithVersion: false,
+			}),
+	)
+	if err != nil {
+		return err
+	}
+
+	c.Client = db
+
+	return nil
+}
+
+const (
+	ShowDatabases = "SHOW DATABASES"
+	ShowTables    = "SELECT name FROM system.tables WHERE database = '%s'"
+	DescTable     = "SELECT name,type FROM system.columns WHERE database='%s' AND table = '%s';"
+)
+
+func (c *Clickhouse) QueryRows(ctx context.Context, query string) (*sql.Rows, error) {
+	var (
+		rows *sql.Rows
+		err  error
+	)
+
+	if c.ClientByHTTP != nil {
+		rows, err = c.ClientByHTTP.Query(query)
+		if err != nil {
+			return nil, err
+		}
+	} else if c.Client != nil {
+		rows, err = c.Client.Raw(query).Rows()
+		if err != nil {
+			return nil, err
+		}
+	} else {
+		return nil, fmt.Errorf("clickhouse client is nil")
+	}
+
+	return rows, nil
+}
+
+// ShowDatabases lists all databases in Clickhouse
+func (c *Clickhouse) ShowDatabases(ctx context.Context) ([]string, error) {
+	var (
+		res []string
+	)
+
+	rows, err := c.QueryRows(ctx, ShowDatabases)
+	if err != nil {
+		return nil, err
+	}
+
+	for rows.Next() {
+		var r string
+		if err := rows.Scan(&r); err != nil {
+			return nil, err
+		}
+		res = append(res, r)
+	}
+
+	return res, nil
+}
+
+// ShowTables lists all tables in a given database
+func (c *Clickhouse) ShowTables(ctx context.Context, database string) ([]string, error) {
+	var (
+		res []string
+	)
+
+	showTables := fmt.Sprintf(ShowTables, database)
+	rows, err := c.QueryRows(ctx, showTables)
+	if err != nil {
+		return nil, err
+	}
+
+	for rows.Next() {
+		var r string
+		if err := rows.Scan(&r); err != nil {
+			return nil, err
+		}
+		res = append(res, r)
+	}
+
+	return res, nil
+}
+
+// DescribeTable describes the schema of a specified table in Clickhouse
+func (c *Clickhouse) DescribeTable(ctx context.Context, query interface{}) ([]*types.ColumnProperty, error) {
+	var (
+		ret []*types.ColumnProperty
+	)
+
+	ckQueryParam := new(QueryParam)
+	if err := mapstructure.Decode(query, ckQueryParam); err != nil {
+		return nil, err
+	}
+	descTable := fmt.Sprintf(DescTable, ckQueryParam.Database, ckQueryParam.Table)
+
+	rows, err := c.QueryRows(ctx, descTable)
+	if err != nil {
+		return nil, err
+	}
+
+	for rows.Next() {
+		var column types.ColumnProperty
+		if err := rows.Scan(&column.Field, &column.Type); err != nil {
+			return nil, err
+		}
+		ret = append(ret, &column)
+	}
+
+	return ret, nil
+}
+
+func (c *Clickhouse) ExecQueryBySqlDB(ctx context.Context, sql string) ([]map[string]interface{}, error) {
+	rows, err := c.QueryRows(ctx, sql)
+
+	if err != nil {
+		return nil, err
+	}
+	defer rows.Close()
+
+	columns, err := rows.Columns()
+	if err != nil {
+		return nil, err
+	}
+
+	var results []map[string]interface{}
+
+	for rows.Next() {
+		columnValues := make([]interface{}, len(columns))
+		columnPointers := make([]interface{}, len(columns))
+		for i := range columnValues {
+			columnPointers[i] = &columnValues[i]
+		}
+
+		if err := rows.Scan(columnPointers...); err != nil {
+			continue
+		}
+
+		rowMap := make(map[string]interface{})
+		for i, colName := range columns {
+			val := columnValues[i]
+			bytes, ok := val.([]byte)
+			if ok {
+				rowMap[colName] = string(bytes)
+			} else {
+				rowMap[colName] = val
+			}
+		}
+		results = append(results, rowMap)
+	}
+	return results, nil
+}
+
+func (c *Clickhouse) Query(ctx context.Context, query interface{}) ([]map[string]interface{}, error) {
+
+	ckQuery := new(QueryParam)
+	if err := mapstructure.Decode(query, ckQuery); err != nil {
+		return nil, err
+	}
+
+	// 校验SQL的合法性, 过滤掉 write请求
+	sqlItem := strings.Split(strings.ToUpper(ckQuery.Sql), " ")
+	for _, item := range sqlItem {
+		if _, ok := ckBannedOp[item]; ok {
+			return nil, fmt.Errorf("operation %s is forbid, only read db, please check your sql", item)
+		}
+	}
+
+	// 检查匹配数据长度，防止数据量过大
+	err := c.CheckMaxQueryRows(ctx, ckQuery.Sql)
+	if err != nil {
+		return nil, err
+	}
+
+	dbRows := make([]map[string]interface{}, 0)
+
+	if c.ClientByHTTP != nil {
+		dbRows, err = c.ExecQueryBySqlDB(ctx, ckQuery.Sql)
+	} else {
+		err = c.Client.Raw(ckQuery.Sql).Find(&dbRows).Error
+	}
+	if err != nil {
+		return nil, fmt.Errorf("fetch data failed, sql is %s, err is %s", ckQuery.Sql, err.Error())
+	}
+
+	return dbRows, nil
+}
+
+func (c *Clickhouse) CheckMaxQueryRows(ctx context.Context, sql string) error {
+
+	subSql := strings.ReplaceAll(sql, ";", "")
+	subSql = fmt.Sprintf("SELECT COUNT(*) as count FROM (%s) AS subquery;", subSql)
+
+	dbRows, err := c.ExecQueryBySqlDB(ctx, subSql)
+	if err != nil {
+		return fmt.Errorf("fetch data failed, sql is %s, err is %s", subSql, err.Error())
+	}
+
+	if len(dbRows) > 0 {
+		if count, exists := dbRows[0]["count"]; exists {
+			v, err := sqlbase.ParseFloat64Value(count)
+			if err != nil {
+				return err
+			}
+
+			if v > float64(c.MaxQueryRows) {
+				return fmt.Errorf("query result rows count %d exceeds the maximum limit %d", int(v), c.MaxQueryRows)
+			}
+		}
+	}
+
+	return nil
+}

dskit/clickhouse/clickhouse_test.go

@@ -0,0 +1,43 @@
+package clickhouse
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/dskit/types"
+)
+
+func Test_Timeseries(t *testing.T) {
+	ck := &Clickhouse{
+		Nodes:    []string{"127.0.0.1:8123"},
+		User:     "default",
+		Password: "123456",
+	}
+
+	err := ck.InitCli()
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	data, err := ck.QueryTimeseries(context.TODO(), &QueryParam{
+		Sql:        `select * from default.student limit 20`,
+		From:       time.Now().Unix() - 300,
+		To:         time.Now().Unix(),
+		TimeField:  "created_at",
+		TimeFormat: "datetime",
+		Keys: types.Keys{
+			LabelKey: "age",
+		},
+	})
+	if err != nil {
+		t.Fatal(err)
+	}
+	bs, err := json.Marshal(data)
+	if err != nil {
+		t.Fatal(err)
+	}
+	fmt.Println(string(bs))
+}

dskit/clickhouse/timeseries.go

@@ -0,0 +1,58 @@
+package clickhouse
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/ccfos/nightingale/v6/dskit/sqlbase"
+	"github.com/ccfos/nightingale/v6/dskit/types"
+)
+
+const (
+	TimeFieldFormatEpochMilli  = "epoch_millis"
+	TimeFieldFormatEpochSecond = "epoch_second"
+)
+
+// 时序数据相关的API
+type QueryParam struct {
+	Limit      int        `json:"limit" mapstructure:"limit"`
+	Sql        string     `json:"sql" mapstructure:"sql"`
+	Ref        string     `json:"ref" mapstructure:"ref"`
+	From       int64      `json:"from" mapstructure:"from"`
+	To         int64      `json:"to" mapstructure:"to"`
+	TimeField  string     `json:"time_field" mapstructure:"time_field"`
+	TimeFormat string     `json:"time_format" mapstructure:"time_format"`
+	Keys       types.Keys `json:"keys" mapstructure:"keys"`
+	Database   string     `json:"database" mapstructure:"database"`
+	Table      string     `json:"table" mapstructure:"table"`
+}
+
+var (
+	ckBannedOp = map[string]struct{}{
+		"CREATE":   {},
+		"INSERT":   {},
+		"ALTER":    {},
+		"REVOKE":   {},
+		"DROP":     {},
+		"RENAME":   {},
+		"ATTACH":   {},
+		"DETACH":   {},
+		"OPTIMIZE": {},
+		"TRUNCATE": {},
+		"SET":      {},
+	}
+)
+
+func (c *Clickhouse) QueryTimeseries(ctx context.Context, query *QueryParam) ([]types.MetricValues, error) {
+	if query.Keys.ValueKey == "" {
+		return nil, fmt.Errorf("valueKey is required")
+	}
+
+	rows, err := c.Query(ctx, query)
+	if err != nil {
+		return nil, err
+	}
+
+	// 构造成时续数据
+	return sqlbase.FormatMetricValues(query.Keys, rows, true), nil
+}

dskit/sqlbase/base.go

@@ -0,0 +1,236 @@
+// @Author: Ciusyan 5/19/24
+
+package sqlbase
+
+import (
+	"context"
+	"database/sql"
+	"fmt"
+	"strings"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/dskit/types"
+
+	"gorm.io/gorm"
+)
+
+// NewDB creates a new Gorm DB instance based on the provided gorm.Dialector and configures the connection pool
+func NewDB(ctx context.Context, dialector gorm.Dialector, maxIdleConns, maxOpenConns int, connMaxLifetime time.Duration) (*gorm.DB, error) {
+	// Create a new Gorm DB instance
+	db, err := gorm.Open(dialector, &gorm.Config{})
+	if err != nil {
+		return nil, err
+	}
+
+	// Configure the connection pool
+	sqlDB, err := db.DB()
+	if err != nil {
+		return nil, err
+	}
+
+	sqlDB.SetMaxIdleConns(maxIdleConns)
+	sqlDB.SetMaxOpenConns(maxOpenConns)
+	sqlDB.SetConnMaxLifetime(connMaxLifetime)
+
+	return db.WithContext(ctx), sqlDB.Ping()
+}
+
+// ShowTables retrieves a list of all tables in the specified database
+func ShowTables(ctx context.Context, db *gorm.DB, query string) ([]string, error) {
+	var tables []string
+
+	rows, err := db.WithContext(ctx).Raw(query).Rows()
+	if err != nil {
+		return nil, err
+	}
+	defer rows.Close()
+
+	for rows.Next() {
+		var table string
+		if err := rows.Scan(&table); err != nil {
+			return nil, err
+		}
+		tables = append(tables, table)
+	}
+
+	return tables, nil
+}
+
+// ShowDatabases retrieves a list of all databases in the connected database server
+func ShowDatabases(ctx context.Context, db *gorm.DB, query string) ([]string, error) {
+	var databases []string
+
+	rows, err := db.WithContext(ctx).Raw(query).Rows()
+	if err != nil {
+		return nil, err
+	}
+	defer rows.Close()
+
+	for rows.Next() {
+		var database string
+		if err := rows.Scan(&database); err != nil {
+			return nil, err
+		}
+		databases = append(databases, database)
+	}
+
+	return databases, nil
+}
+
+// DescTable describes the schema of a specified table in MySQL or PostgreSQL
+func DescTable(ctx context.Context, db *gorm.DB, query string) ([]*types.ColumnProperty, error) {
+	rows, err := db.WithContext(ctx).Raw(query).Rows()
+	if err != nil {
+		return nil, err
+	}
+	defer rows.Close()
+
+	var columns []*types.ColumnProperty
+	for rows.Next() {
+		var (
+			field        string
+			typ          string
+			null         string
+			key          sql.NullString
+			defaultValue sql.NullString
+			extra        sql.NullString
+		)
+
+		switch db.Dialector.Name() {
+		case "mysql":
+			if err := rows.Scan(&field, &typ, &null, &key, &defaultValue, &extra); err != nil {
+				continue
+			}
+		case "postgres", "sqlserver":
+			if err := rows.Scan(&field, &typ, &null, &defaultValue); err != nil {
+				continue
+			}
+		case "oracle":
+			if err := rows.Scan(&field, &typ, &null); err != nil {
+				continue
+			}
+		}
+
+		// Convert the database-specific type to internal type
+		type2, indexable := convertDBType(db.Dialector.Name(), typ)
+		columns = append(columns, &types.ColumnProperty{
+			Field:     field,
+			Type:      typ,
+			Type2:     type2,
+			Indexable: indexable,
+		})
+	}
+	return columns, nil
+}
+
+// ExecQuery executes the specified query and returns the result rows
+func ExecQuery(ctx context.Context, db *gorm.DB, sql string) ([]map[string]interface{}, error) {
+	rows, err := db.WithContext(ctx).Raw(sql).Rows()
+	if err != nil {
+		return nil, err
+	}
+	defer rows.Close()
+
+	columns, err := rows.Columns()
+	if err != nil {
+		return nil, err
+	}
+
+	var results []map[string]interface{}
+
+	for rows.Next() {
+		columnValues := make([]interface{}, len(columns))
+		columnPointers := make([]interface{}, len(columns))
+		for i := range columnValues {
+			columnPointers[i] = &columnValues[i]
+		}
+
+		if err := rows.Scan(columnPointers...); err != nil {
+			continue
+		}
+
+		rowMap := make(map[string]interface{})
+		for i, colName := range columns {
+			val := columnValues[i]
+			bytes, ok := val.([]byte)
+			if ok {
+				rowMap[colName] = string(bytes)
+			} else {
+				rowMap[colName] = val
+			}
+		}
+		results = append(results, rowMap)
+	}
+	return results, nil
+}
+
+// SelectRows selects rows from a specified table based on a given query
+func SelectRows(ctx context.Context, db *gorm.DB, table, query string) ([]map[string]interface{}, error) {
+	sql := fmt.Sprintf("SELECT * FROM %s", table)
+	if query != "" {
+		sql += " WHERE " + query
+	}
+
+	return ExecQuery(ctx, db, sql)
+}
+
+// convertDBType converts MySQL or PostgreSQL data types to custom internal types and determines if they are indexable
+func convertDBType(dialect, dbType string) (string, bool) {
+	typ := strings.ToLower(dbType)
+
+	// Common type conversions
+	switch {
+	case strings.HasPrefix(typ, "int"), strings.HasPrefix(typ, "tinyint"),
+		strings.HasPrefix(typ, "smallint"), strings.HasPrefix(typ, "mediumint"),
+		strings.HasPrefix(typ, "bigint"), strings.HasPrefix(typ, "serial"),
+		strings.HasPrefix(typ, "bigserial"):
+		return types.LogExtractValueTypeLong, true
+
+	case strings.HasPrefix(typ, "varchar"), strings.HasPrefix(typ, "text"),
+		strings.HasPrefix(typ, "char"), strings.HasPrefix(typ, "tinytext"),
+		strings.HasPrefix(typ, "mediumtext"), strings.HasPrefix(typ, "longtext"),
+		strings.HasPrefix(typ, "character varying"), strings.HasPrefix(typ, "nvarchar"),
+		strings.HasPrefix(typ, "nchar"):
+		return types.LogExtractValueTypeText, true
+
+	case strings.HasPrefix(typ, "float"), strings.HasPrefix(typ, "double"),
+		strings.HasPrefix(typ, "decimal"), strings.HasPrefix(typ, "numeric"),
+		strings.HasPrefix(typ, "real"), strings.HasPrefix(typ, "double precision"):
+		return types.LogExtractValueTypeFloat, true
+
+	case strings.HasPrefix(typ, "date"), strings.HasPrefix(typ, "datetime"),
+		strings.HasPrefix(typ, "timestamp"), strings.HasPrefix(typ, "timestamptz"),
+		strings.HasPrefix(typ, "time"), strings.HasPrefix(typ, "smalldatetime"):
+		return types.LogExtractValueTypeDate, false
+
+	case strings.HasPrefix(typ, "boolean"), strings.HasPrefix(typ, "bit"):
+		return types.LogExtractValueTypeBool, false
+	}
+
+	// Specific type conversions for MySQL
+	if dialect == "mysql" {
+		switch {
+		default:
+			return typ, false
+		}
+	}
+
+	// Specific type conversions for PostgreSQL
+	if dialect == "postgres" {
+		switch {
+		default:
+			return typ, false
+		}
+	}
+
+	if dialect == "oracle" {
+		switch {
+		default:
+			return typ, false
+		}
+	}
+
+	// Can continue to add specific 'dialect' type ...
+
+	return typ, false
+}

dskit/sqlbase/timeseries.go

@@ -0,0 +1,250 @@
+// @Author: Ciusyan 5/20/24
+
+package sqlbase
+
+import (
+	"context"
+	"crypto/md5"
+	"encoding/json"
+	"fmt"
+	"reflect"
+	"sort"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/dskit/types"
+
+	"github.com/prometheus/common/model"
+	"gorm.io/gorm"
+)
+
+type QueryParam struct {
+	Sql  string     `json:"sql"`
+	Keys types.Keys `json:"keys" mapstructure:"keys"`
+}
+
+var (
+	BannedOp = map[string]struct{}{
+		"CREATE":   {},
+		"INSERT":   {},
+		"UPDATE":   {},
+		"DELETE":   {},
+		"ALTER":    {},
+		"REVOKE":   {},
+		"DROP":     {},
+		"RENAME":   {},
+		"TRUNCATE": {},
+		"SET":      {},
+	}
+)
+
+// Query executes a given SQL query and returns the results
+func Query(ctx context.Context, db *gorm.DB, query *QueryParam) ([]map[string]interface{}, error) {
+	// Validate SQL to prevent write operations if needed
+	sqlItem := strings.Split(strings.ToUpper(query.Sql), " ")
+	for _, item := range sqlItem {
+		if _, ok := BannedOp[item]; ok {
+			return nil, fmt.Errorf("operation %s is forbidden, only read operations are allowed, please check your SQL", item)
+		}
+	}
+
+	return ExecQuery(ctx, db, query.Sql)
+}
+
+// QueryTimeseries executes a time series data query using the given parameters
+func QueryTimeseries(ctx context.Context, db *gorm.DB, query *QueryParam, ignoreDefault ...bool) ([]types.MetricValues, error) {
+	rows, err := Query(ctx, db, query)
+	if err != nil {
+		return nil, err
+	}
+
+	return FormatMetricValues(query.Keys, rows, ignoreDefault...), nil
+}
+
+func FormatMetricValues(keys types.Keys, rows []map[string]interface{}, ignoreDefault ...bool) []types.MetricValues {
+	ignore := false
+	if len(ignoreDefault) > 0 {
+		ignore = ignoreDefault[0]
+	}
+
+	keyMap := make(map[string]string)
+	for _, valueMetric := range strings.Split(keys.ValueKey, " ") {
+		keyMap[valueMetric] = "value"
+	}
+
+	for _, labelMetric := range strings.Split(keys.LabelKey, " ") {
+		keyMap[labelMetric] = "label"
+	}
+
+	if keys.TimeKey == "" {
+		keys.TimeKey = "time"
+	}
+
+	if len(keys.TimeKey) > 0 {
+		keyMap[keys.TimeKey] = "time"
+	}
+
+	var dataResps []types.MetricValues
+	dataMap := make(map[string]*types.MetricValues)
+
+	for _, row := range rows {
+		labels := make(map[string]string)
+		metricValue := make(map[string]float64)
+		metricTs := make(map[string]float64)
+
+		// Process each column based on its designated role (value, label, time)
+		for k, v := range row {
+			switch keyMap[k] {
+			case "value":
+				val, err := ParseFloat64Value(v)
+				if err != nil {
+					continue
+				}
+				metricValue[k] = val
+			case "label":
+				labels[k] = fmt.Sprintf("%v", v)
+			case "time":
+				ts, err := ParseTime(v, keys.TimeFormat)
+				if err != nil {
+					continue
+				}
+				metricTs[k] = float64(ts.Unix())
+			default:
+				// Default to labels for any unrecognized columns
+				if !ignore {
+					labels[k] = fmt.Sprintf("%v", v)
+				}
+			}
+		}
+
+		// Compile and store the metric values
+		for metricName, value := range metricValue {
+			metrics := make(model.Metric)
+			var labelsStr []string
+
+			for k1, v1 := range labels {
+				metrics[model.LabelName(k1)] = model.LabelValue(v1)
+				labelsStr = append(labelsStr, fmt.Sprintf("%s=%s", k1, v1))
+			}
+			metrics["__name__"] = model.LabelValue(metricName)
+			labelsStr = append(labelsStr, fmt.Sprintf("__name__=%s", metricName))
+
+			// Hash the labels to use as a key
+			sort.Strings(labelsStr)
+			labelsStrHash := fmt.Sprintf("%x", md5.Sum([]byte(strings.Join(labelsStr, ","))))
+
+			// Append new values to the existing metric, if present
+			ts, exists := metricTs[keys.TimeKey]
+			if !exists {
+				ts = float64(time.Now().Unix()) // Default to current time if not specified
+			}
+
+			valuePair := []float64{ts, value}
+			if existing, ok := dataMap[labelsStrHash]; ok {
+				existing.Values = append(existing.Values, valuePair)
+			} else {
+				dataResp := types.MetricValues{
+					Metric: metrics,
+					Values: [][]float64{valuePair},
+				}
+				dataMap[labelsStrHash] = &dataResp
+			}
+		}
+	}
+
+	// Convert the map to a slice for the response
+	for _, v := range dataMap {
+		sort.Slice(v.Values, func(i, j int) bool { return v.Values[i][0] < v.Values[j][0] }) // Sort by timestamp
+		dataResps = append(dataResps, *v)
+	}
+
+	return dataResps
+}
+
+// parseFloat64Value attempts to convert an interface{} to float64 using reflection
+func ParseFloat64Value(val interface{}) (float64, error) {
+	v := reflect.ValueOf(val)
+	switch v.Kind() {
+	case reflect.Float64, reflect.Float32:
+		return v.Float(), nil
+	case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
+		return float64(v.Int()), nil
+	case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
+		return float64(v.Uint()), nil
+	case reflect.String:
+		return strconv.ParseFloat(v.String(), 64)
+	case reflect.Slice:
+		if v.Type().Elem().Kind() == reflect.Uint8 {
+			return strconv.ParseFloat(string(v.Bytes()), 64)
+		}
+	case reflect.Interface:
+		return ParseFloat64Value(v.Interface())
+	case reflect.Ptr:
+		if !v.IsNil() {
+			return ParseFloat64Value(v.Elem().Interface())
+		}
+	case reflect.Struct:
+		if num, ok := val.(json.Number); ok {
+			return num.Float64()
+		}
+	}
+	return 0, fmt.Errorf("cannot convert type %T to float64", val)
+}
+
+// ParseTime attempts to parse a time value from an interface{} using a specified format
+func ParseTime(val interface{}, format string) (time.Time, error) {
+	v := reflect.ValueOf(val)
+	switch v.Kind() {
+	case reflect.String:
+		str := v.String()
+		return parseTimeFromString(str, format)
+	case reflect.Slice:
+		if v.Type().Elem().Kind() == reflect.Uint8 {
+			str := string(v.Bytes())
+			return parseTimeFromString(str, format)
+		}
+	case reflect.Int, reflect.Int64:
+		return time.Unix(v.Int(), 0), nil
+	case reflect.Float64:
+		return time.Unix(int64(v.Float()), 0), nil
+	case reflect.Interface:
+		return ParseTime(v.Interface(), format)
+	case reflect.Ptr:
+		if !v.IsNil() {
+			return ParseTime(v.Elem().Interface(), format)
+		}
+	case reflect.Struct:
+		if t, ok := val.(time.Time); ok {
+			return t, nil
+		}
+	}
+	return time.Time{}, fmt.Errorf("invalid time value type: %v", val)
+}
+
+func parseTimeFromString(str, format string) (time.Time, error) {
+	// If a custom time format is provided, use it to parse the string
+	if format != "" {
+		parsedTime, err := time.Parse(format, str)
+		if err == nil {
+			return parsedTime, nil
+		}
+		return time.Time{}, fmt.Errorf("failed to parse time '%s' with format '%s': %v", str, format, err)
+	}
+
+	// Try to parse the string as RFC3339, RFC3339Nano, or Unix timestamp
+	if parsedTime, err := time.Parse(time.RFC3339, str); err == nil {
+		return parsedTime, nil
+	}
+	if parsedTime, err := time.Parse(time.RFC3339Nano, str); err == nil {
+		return parsedTime, nil
+	}
+	if timestamp, err := strconv.ParseInt(str, 10, 64); err == nil {
+		return time.Unix(timestamp, 0), nil
+	}
+	if timestamp, err := strconv.ParseFloat(str, 64); err == nil {
+		return time.Unix(int64(timestamp), 0), nil
+	}
+
+	return time.Time{}, fmt.Errorf("failed to parse time '%s'", str)
+}

dskit/sqlbase/timeseries_test.go

@@ -0,0 +1,149 @@
+// @Author: Ciusyan 5/17/24
+
+package sqlbase
+
+import (
+	"encoding/json"
+	"testing"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/dskit/types"
+)
+
+func TestFormatMetricValues(t *testing.T) {
+	tests := []struct {
+		name string
+		keys types.Keys
+		rows []map[string]interface{}
+		want []types.MetricValues
+	}{
+		{
+			name: "cases1",
+			keys: types.Keys{
+				ValueKey:   "grade a_grade",
+				LabelKey:   "id student_name",
+				TimeKey:    "update_time",
+				TimeFormat: "2006-01-02 15:04:05",
+			},
+			rows: []map[string]interface{}{
+				{
+					"id":           "10007",
+					"grade":        20003,
+					"student_name": "邵子韬",
+					"a_grade":      69,
+					"update_time":  "2024-05-14 10:00:00",
+				},
+				{
+					"id":           "10007",
+					"grade":        20003,
+					"student_name": "邵子韬",
+					"a_grade":      69,
+					"update_time":  "2024-05-14 10:05:00",
+				},
+				{
+					"id":           "10007",
+					"grade":        20003,
+					"student_name": "邵子韬",
+					"a_grade":      69,
+					"update_time":  "2024-05-14 10:10:00",
+				},
+				{
+					"id":           "10008",
+					"grade":        20004,
+					"student_name": "Ciusyan",
+					"a_grade":      100,
+					"update_time":  "2024-05-14 12:00:00",
+				},
+			},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := FormatMetricValues(tt.keys, tt.rows)
+			for _, g := range got {
+				t.Log(g)
+			}
+		})
+	}
+}
+
+func TestParseFloat64Value(t *testing.T) {
+
+	ptr := func(val float64) *float64 {
+		return &val
+	}
+
+	tests := []struct {
+		name    string
+		input   interface{}
+		want    float64
+		wantErr bool
+	}{
+		{"float64", 1.23, 1.23, false},
+		{"float32", float32(1.23), float64(float32(1.23)), false},
+		{"int", 123, 123, false},
+		{"int64", int64(123), 123, false},
+		{"uint", uint(123), 123, false},
+		{"uint64", uint64(123), 123, false},
+		{"string", "1.23", 1.23, false},
+		{"[]byte", []byte("1.23"), 1.23, false},
+		{"json.Number", json.Number("1.23"), 1.23, false},
+		{"interface", interface{}(1.23), 1.23, false},
+		{"pointer", ptr(1.23), 1.23, false},
+		{"invalid string", "abc", 0, true},
+		{"invalid type", struct{}{}, 0, true},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := ParseFloat64Value(tt.input)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("parseFloat64Value() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if got != tt.want {
+				t.Errorf("parseFloat64Value() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func TestParseTime(t *testing.T) {
+
+	ptrTime := func(t time.Time) *time.Time {
+		return &t
+	}
+
+	tests := []struct {
+		name    string
+		input   interface{}
+		format  string
+		want    time.Time
+		wantErr bool
+	}{
+		{"RFC3339", "2024-05-14T12:34:56Z", "", time.Date(2024, 5, 14, 12, 34, 56, 0, time.UTC), false},
+		{"RFC3339Nano", "2024-05-14T12:34:56.789Z", "", time.Date(2024, 5, 14, 12, 34, 56, 789000000, time.UTC), false},
+		{"Unix timestamp int", int64(1715642135), "", time.Unix(1715642135, 0), false},
+		{"Unix timestamp float64", 1715642135.0, "", time.Unix(int64(1715642135), 0), false},
+		{"custom format", "14/05/2024", "02/01/2006", time.Date(2024, 5, 14, 0, 0, 0, 0, time.UTC), false},
+		{"slice", []byte("2024-05-14T12:34:56Z"), "", time.Date(2024, 5, 14, 12, 34, 56, 0, time.UTC), false},
+		{"interface", interface{}("2024-05-14T12:34:56Z"), "", time.Date(2024, 5, 14, 12, 34, 56, 0, time.UTC), false},
+		{"pointer", ptrTime(time.Date(2024, 5, 14, 12, 34, 56, 0, time.UTC)), "", time.Date(2024, 5, 14, 12, 34, 56, 0, time.UTC), false},
+		{"invalid format", "14-05-2024", "02/01/2006", time.Time{}, true},
+		{"invalid type", struct{}{}, "", time.Time{}, true},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := ParseTime(tt.input, tt.format)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("ParseTime() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !got.Equal(tt.want) {
+				t.Errorf("ParseTime() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}