Compare commits

..

123 Commits

Author SHA1 Message Date
Jeff McCune
bd56f3118c v0.96.0 - v1alpha4 Core API
Release v0.96.0 with the new Core API design.
2024-10-10 13:55:06 -07:00
Jeff McCune
d3aa748e92 v1alpha4: add file generator (#277)
Previously the file generator was unimplemented.  This patch implements
it as a simple file read into the ArtifactMap for use by the Kustomize
or Join transformers.

With this patch all v1alpha4 Core API features are implemented.
Resources, Helm, and File generators.  Kustomize and Join transformers.
2024-10-10 13:52:08 -07:00
Jeff McCune
6738248756 v1alpha4: fix blank log lines
Blank lines show up in the output which is confusing. This patch fixes
the only source location identified with the following command.

    export HOLOS_LOG_LEVEL=debug
    export HOLOS_LOG_FORMAT=json
    holos render platform ./platform 2>&1 | jq -r 'select (.msg == "")'
2024-10-10 13:44:08 -07:00
Jeff McCune
011b488775 v1alpha4: cache helm charts by version (#273)
Previously helm charts were cached only by name, which is a problem
because the wrong version would be used when previously cached.

This patch caches charts by name and version to ensure changes in the
version results in pulling the new cached version.  It is the user's
responsibility to remove old versions.

This patch also ensures only a single go routine can run cacheChart() at
a time across processes.  This is necessary when rendering a platform
because multiple processes will run the Helm generator concurrently, for
example when the same chart is used for multiple environments or
customers.

The mkdir system call serves as the locking mechanism, which is robust
and atomic on all commonly used file systems.
2024-10-10 12:15:20 -07:00
Jeff McCune
c8d89f3291 v1alpha4: add helm chart generator
Previously the helm generator was not implemented and returned an error.
This patch is a first pass copying the helm method from
internal/render/helm.go

Basic testing performed with a podinfo chart.  It works as the previous
versions in v1alpha3 and before works.  This patch does not address the
cached version issue in #273
2024-10-10 08:25:13 -07:00
Jeff McCune
a44ebe5171 refactor Artifact to use strings instead of FilePath
holos.FilePath is intended for paths relative to the platform root
directory.  We use the Artifact to store lots of stuff not related to
the platform root directory, for example kustomization.yaml in a temp
dir.  Most entries are not relative to the platform root directory given
the implicit cfg.WriteTo prefix.
2024-10-09 20:16:13 -07:00
Jeff McCune
66a3b6a874 improve error of missing key
Previously:

  could not run: could not build dev-join: could not get foo.yaml: not set at builder/v1alpha4/builder.go:180

This is confusing because set has nothing to do with the missing input
from the cue code the user writes.

Result:

  could not run: could not build test-join: missing foo.yaml at builder/v1alpha4/builder.go:180

This is better because it at doesn't distract the user from the fact
they're missing a foo.yaml generator output to align with the
transformer input.
2024-10-09 19:39:33 -07:00
Jeff McCune
0886788238 Merge pull request #272 from holos-run/jeff/268-v1alpha4
v1alpha4: add join transformer
2024-10-09 17:45:28 -07:00
Jeff McCune
4fd6785a10 v1alpha4: add Join transformer
The Join transformer was not implemented.  This patch completes the
transformers we support, the Join and Kustomize transformers.
2024-10-09 17:42:56 -07:00
Jeff McCune
e4695fa204 artifact: add Save() method to save artifacts
The code was inlined in a number of places, it makes sense to move it to
the interface.  It'll also make it easier to test, we can provide a null
writer concrete value.
2024-10-09 16:53:31 -07:00
Jeff McCune
4cd9395e6c v1alpha4: add concurrent build plan artifacts
Previously the Artifact collection was processed sequentially.  This
patch provides a modest performance improvement, about 16% faster for
our simple 2 artifact use case, by processing each artifact
concurrently.
2024-10-09 16:15:46 -07:00
Jeff McCune
6f4f355ee0 Merge pull request #270 from holos-run/jeff/268-v1alpha4
v1alpha4 - Resources and Kustomize
2024-10-09 15:28:13 -07:00
Jeff McCune
0fde16f477 v1alpha4: fix lint 2024-10-09 14:43:03 -07:00
Jeff McCune
426b4323f7 v1alpha4: inject component build plan name from platform
Platform rendering provides poor user feedback:

```
❯ holos render platform ./platforms/minimal
rendered namespaces for cluster local in 143.068583ms
rendered namespaces for cluster local in 143.861834ms
rendered namespaces for cluster local in 144.072666ms
rendered namespaces for cluster local in 144.219417ms
rendered platform in 144.326625ms
```

We want to see the metadata.name field of each BuildPlan.  This patch
injects the build plan name from the platform spec to make the name
available through the end to end platform rendering process.

Result:

```
❯ holos render platform ./platforms/minimal
rendered stage-namespaces for cluster local in 146.078375ms
rendered prod-namespaces for cluster local in 146.544583ms
rendered test-namespaces for cluster local in 147.0535ms
rendered dev-namespaces for cluster local in 147.499166ms
rendered platform in 147.553875ms
```
2024-10-09 14:06:42 -07:00
Jeff McCune
ee1e4988a6 v1alpha4: write fully rendered build plan artifacts
With this patch the first use case of CUE Resources + Kustomize is fully
working, artifacts are written into the deploy directory.

❯ holos render platform ./platforms/minimal
rendered namespaces for cluster local in 143.068583ms
rendered namespaces for cluster local in 143.861834ms
rendered namespaces for cluster local in 144.072666ms
rendered namespaces for cluster local in 144.219417ms
rendered platform in 144.326625ms

The output indicates we need to plumb the BuildPlan metadata.name from
the PlatfromSpec through to the render component command.  This is
necessary so we can report the correct name instead of just the base
path.
2024-10-09 14:06:42 -07:00
Jeff McCune
5c391e8444 v1alpha4: refactor error handing for resources and kustomize
Consistently use errors.Format("%s: %w", msg, err)
2024-10-09 14:06:41 -07:00
Jeff McCune
aba1b44f4d v1alpha4: fix resources manifest generation
Without this patch holos writes a single yaml document that is a list.
It needs to write a file that contains multiple documents, each document
a map[string]any representing the kubernetes resource.

This patch fixes the problem.  With this patch kustomize fully executes.
2024-10-09 14:06:41 -07:00
Jeff McCune
44f9615a93 v1alpha4: implement kustomize transformer 2024-10-09 14:06:41 -07:00
Jeff McCune
69a6d2acad v1alpha4: implement resources generator 2024-10-09 14:06:40 -07:00
Jeff McCune
d2c94dc8df refactor artifact manifest field to inputs and output
The manifest field isn't clear.

Much more clear to have generators produce one Output.  Transformers
take multiple Inputs and produce one Output.

The final Transformer, or a single Generator, must produce the final
Artifact.

The Inputs and Output naming to produce an Artifact makes clear the
rendering pipeline we're implementing.

This also makes clear that multiple generators must have at least one
transformer to produce the final output artifact.  We model a simple
Join transformer for this case, which is what `holos` was implicitly
doing previously.
2024-10-09 14:06:40 -07:00
Jeff McCune
01720b38fc refactor BuildContext back to Component
Component makes much more sense, that's the domain terminology we use.
BuildContext was meant to be re-used elsewhere, but we never did so the
name serves no purpose.
2024-10-09 14:06:40 -07:00
Jeff McCune
632e3c2725 refactor BuildStep to Artifact
Step doesn't make sense, they're not sequential.  They're meant to be
concurrently produced, and the whole point is to produce one artifact so
we might as well call it an Artifact.

```
kind: BuildPlan
apiVersion: v1alpha4
metadata:
  name: prod-namespaces
spec:
  component: projects/platform/components/namespaces
  artifacts:
    - artifact: clusters/no-cluster/components/prod-namespaces/prod-namespaces.gen.yaml
      generators:
        - kind: Resources
          manifest: resources.gen.yaml
          resources:
            Namespace:
              prod-jeff:
                metadata:
                  name: prod-jeff
                  labels:
                    kubernetes.io/metadata.name: prod-jeff
                kind: Namespace
                apiVersion: v1
              prod-gary:
                metadata:
                  name: prod-gary
                  labels:
                    kubernetes.io/metadata.name: prod-gary
                kind: Namespace
                apiVersion: v1
              prod-nate:
                metadata:
                  name: prod-nate
                  labels:
                    kubernetes.io/metadata.name: prod-nate
                kind: Namespace
                apiVersion: v1
      transformers:
        - kind: Kustomize
          kustomize:
            kustomization:
              commonLabels:
                holos.run/component.name: prod-namespaces
              resources:
                - resources.gen.yaml
                - application.gen.yaml
    - artifact: clusters/no-cluster/gitops/prod-namespaces.gen.yaml
      generators:
        - kind: Resources
          manifest: application.gen.yaml
          resources:
            Application:
              argocd:
                apiVersion: argoproj.io/v1alpha1
                kind: Application
                metadata:
                  name: prod-namespaces
                  namespace: argocd
                spec:
                  destination:
                    server: https://kubernetes.default.svc
                  project: default
                  source:
                    path: examples/v1alpha4/deploy/clusters/no-cluster/components/prod-namespaces
                    repoURL: https://github.com/holos-run/bank-of-holos
                    targetRevision: main
      transformers:
        - kind: Kustomize
          kustomize:
            kustomization:
              commonLabels:
                holos.run/component.name: prod-namespaces
              resources:
                - resources.gen.yaml
                - application.gen.yaml
```
2024-10-09 14:06:39 -07:00
Jeff McCune
18e0b48012 refactor generators and transformers
The repeated enabled booleans and file fields are awkward.  It's clear
it's three separate things smashed into one.

kustomize isn't really a generator.  It's useless because there is no
way to reference a plain file in a component directory.

This patch replaces the kustomize generator with a file generator which
simply reads one single file.  Multiple of these generators may be used
to read one or more files.

Then, kustomize may transform these generated files, which are generated
by simply reading from the filesystem.

This API is much improved over the previous.

```
kind: BuildPlan
apiVersion: v1alpha4
metadata:
  name: prod-namespaces
spec:
  component: projects/platform/components/namespaces
  steps:
    - artifact: clusters/no-cluster/components/prod-namespaces/prod-namespaces.gen.yaml
      generators:
        - kind: Resources
          manifest: resources.gen.yaml
          resources:
            Namespace:
              prod-jeff:
                metadata:
                  name: prod-jeff
                  labels:
                    kubernetes.io/metadata.name: prod-jeff
                kind: Namespace
                apiVersion: v1
              prod-gary:
                metadata:
                  name: prod-gary
                  labels:
                    kubernetes.io/metadata.name: prod-gary
                kind: Namespace
                apiVersion: v1
              prod-nate:
                metadata:
                  name: prod-nate
                  labels:
                    kubernetes.io/metadata.name: prod-nate
                kind: Namespace
                apiVersion: v1
      transformers:
        - kind: Kustomize
          kustomize:
            kustomization:
              commonLabels:
                holos.run/component.name: prod-namespaces
              resources:
                - resources.gen.yaml
                - application.gen.yaml
    - artifact: clusters/no-cluster/gitops/prod-namespaces.gen.yaml
      generators:
        - kind: Resources
          manifest: application.gen.yaml
          resources:
            Application:
              argocd:
                apiVersion: argoproj.io/v1alpha1
                kind: Application
                metadata:
                  name: prod-namespaces
                  namespace: argocd
                spec:
                  destination:
                    server: https://kubernetes.default.svc
                  project: default
                  source:
                    path: examples/v1alpha4/deploy/clusters/no-cluster/components/prod-namespaces
                    repoURL: https://github.com/holos-run/bank-of-holos
                    targetRevision: main
      transformers:
        - kind: Kustomize
          kustomize:
            kustomization:
              commonLabels:
                holos.run/component.name: prod-namespaces
              resources:
                - resources.gen.yaml
                - application.gen.yaml
```
2024-10-09 14:06:39 -07:00
Jeff McCune
7dfa9dcb93 remove unnecessary buildstep name
The manifest field is the critical piece of information, not an
arbitrary name.
2024-10-09 14:06:17 -07:00
Jeff McCune
3dedd857ea add manifest field to each build step
A build step either produces kubernetes objects or a gitops manifest.
Both are effectively the same, they're just kubernetes resources.

For the use case of applying common labels to both, we'll have the
Author API pass the same Kustomization to two separate build steps.  One
step to produce the resources, a second to produce the argocd
application or flux kustomization.
2024-10-09 14:06:17 -07:00
Jeff McCune
cd379167cc don't provide a default value for files
Leave this for the author api to avoid always including them in the
exported output.
2024-10-09 14:06:16 -07:00
Jeff McCune
52a5348f82 add build plan build step name
Each step produces a manifest and a gitops file, so we need a unique
name for each step.  The most common case will be a single build step
matching the name of the build plan itself.
2024-10-09 14:06:16 -07:00
Jeff McCune
7c5c8fe692 add fields to store generator output and transformer input
The kustomize transformer needs a filename to store the output from
generators so it has an input for the transformer.  This patch adds
fields for each kind of generator so the kustomize.#Kustomization can be
configured with the files `holos` will write generated output to.
2024-10-09 14:06:15 -07:00
Jeff McCune
ec371ed688 inject missing component tag from render platform to render component 2024-10-09 14:06:15 -07:00
Jeff McCune
1984410577 lift up component path to build plan spec
Hard to use when it's deep inside each build step.
2024-10-09 14:06:15 -07:00
Jeff McCune
438e01fbad replace structpb.Struct with map[string]any
So we can decode using a plain json decoder without needing to marshal
the api objects into yaml from CUE.
2024-10-09 14:06:14 -07:00
Jeff McCune
d122cadae6 render: draft 1 of v1alpha4 core api (#268)
First draft, now to try and wire it up to a platform and namespaces
component.
2024-10-09 14:06:14 -07:00
Jeff McCune
53fcdf307b render: build plan builder (#268)
This patch implements the v1alpha4 component rendering builder for a
component BuildPlan.  We don't yet have the CUE definitions, so this
hasn't been end to end tested yet, the next step is defining the
generators and transforms in the core API BuildPlan.
2024-10-08 06:15:15 -07:00
Jeff McCune
5e6bb96147 render: plumb v1alpha4 component (#268)
This patch plumbs the switch statement to branch on a v1alpha4
BuildPlan.  Tags need to be passed from the render platform subcommand
to the render component subcommand via the --tags argument.
2024-10-07 17:23:21 -07:00
Jeff McCune
94d03f9c59 render: render v1alpha4 platform (#268)
This patch implements minimal rendering of a v1alpha4 platform using the
new render.Builder interface.

Tags aren't wired up yet, but this patch does cleanly separate Builder
interface from the Artifacts.  Platform rendering doesn't have an
artifact itself, all artifacts are produced by rendering each component,
so we'll see how that works when we make the same changes to component
rendering, breaking it down to a render.Builder interface that sets
values in an Artifact.
2024-10-07 16:09:06 -07:00
Jeff McCune
0c01c9177d render: switch platform on api version (#268)
The holos cli does not use an interface to handle different Platform api
versions.  This makes it difficult to evolve the API in a backwards
compatible way.

This patch adds a top level switch statement to the `holos render
platform` command.  The switch discriminates on the Platform API
version.  v1alpha3 and earlier are classified as legacy versions and
will use the existing strict types.  v1alpha4 and later versions will
use an interface to render the platform, allowing for multiple types to
implement the platform rendering interface.
2024-10-07 16:08:59 -07:00
Gary Larizza
36f542da7a Merge pull request #269 from holos-run/gl/landing-page-condense
Make the landing page more concise
2024-10-07 14:17:08 -07:00
Gary Larizza
e41f0aa70c Make the landing page more concise
PROBLEM:

The landing page contains a lot of text, and much of that text was
written before we refined our messaging within the guides and technical
overview pages.

SOLUTION:

* Whittle down landing page text to only the key messages we want to convey.
* Provide messaging bullets for the features.
* Steer folks (via links) to the quickstart guide or technical overview document.

OUTCOME:

Visitors don't need to wade through a lot of text to receive key
messaging talking points or links to the pages they should read.
2024-10-07 11:17:21 -07:00
Jeff McCune
351c8ba74a docs: landing page learn more button to technical overview
Instead of the introduction.
2024-10-04 13:15:28 -07:00
Gary Larizza
a0e0b5bb75 Merge pull request #267 from holos-run/gl/diagram-landing-page
Add Holos diagram to landing page
2024-10-04 12:44:23 -07:00
Gary Larizza
7bc94e314c Add Holos diagram to landing page
PROBLEM:

There's a lot of text to grok on the landing page. A diagram would help
to visually convey what Holos does.

SOLUTION:

* Create a diagram
* Add to landing page

OUTCOME:

A visual aide is present on the landing page that helps explain where
Holos sits.
2024-10-04 12:25:02 -07:00
Jeff McCune
9681ce02e7 docs: clean up tech overview pass 3 (#263) 2024-10-04 08:56:44 -07:00
Jeff McCune
07e8e9f5f3 docs: clean up tech overview pass 2 (#263) 2024-10-03 16:09:13 -07:00
Jeff McCune
437d8a7824 docs: clean up tech overview pass 1 (#263) 2024-10-03 09:56:43 -07:00
Jeff McCune
6cc8214636 docs: fix spelling in technical overview (#263) 2024-10-02 11:29:14 -07:00
Jeff McCune
7d8f324014 blog: add holos technical overview article (#263) 2024-10-02 11:22:08 -07:00
Jeff McCune
8555d56f8c website: enable gtag 2024-09-26 11:55:23 -07:00
Jeff McCune
5884d720f2 website: add mailing lists to footer 2024-09-26 11:38:58 -07:00
Jeff McCune
92d274ced1 website: add support page 2024-09-26 11:30:50 -07:00
Jeff McCune
4dd78bd826 guides: add rendering pipeline diagram to quickstart 2024-09-24 15:24:34 -07:00
Jeff McCune
aeb8fd8e72 blog: add launch article 2024-09-24 15:05:16 -07:00
Gary Larizza
1e9744f748 website: update quickstart narrative (#260)
PROBLEM:

The Quickstart is lacking narrative tying the changes we're asking
people to make to the underlying organizational problems.

SOLUTION:

Improve the narrative to surface the problems we are solving and how
this affects the different teams at the Bank of Holos

OUTCOME:

Clarity on the problems the quickstart is solving.

Closes: #259
2024-09-24 11:37:24 -07:00
Jeff McCune
a95abe65f6 website: refine introduction to be more concise 2024-09-24 07:10:41 -07:00
Jeff McCune
c58510b92c website: refine landing to be more concise 2024-09-24 06:59:14 -07:00
Jeff McCune
302a7bfcf0 guides: fix grammar and tweak the change a service guide 2024-09-23 20:48:35 -07:00
Jeff McCune
decbbaab0f guides: fix typos in change a service guide 2024-09-23 20:04:58 -07:00
Jeff McCune
822f599202 guides: move reset cluster step in change a service
Move to the correct place, after forking the repo.
2024-09-23 20:02:03 -07:00
Jeff McCune
6a47edbc3d guides: Add Change Service Guide (#253)
This patch also adds Organization to the Author API as an example of a
regular expression constraint.
2024-09-23 19:54:46 -07:00
Jeff McCune
67d00f1dd4 website: get rid of What is Holos? on landing page 2024-09-23 14:08:09 -07:00
Jeff McCune
bfa02cd6ed website: introduction (#258) 2024-09-23 13:45:28 -07:00
Jeff McCune
2d8ca474f3 website: Landing Page 3 (#256) 2024-09-23 10:08:25 -07:00
Jeff McCune
c7cd6f5190 website: Landing Page 2 (#256) 2024-09-20 17:14:34 -07:00
Jeff McCune
21e3e6f5e4 website: Landing Page 1 (#256) 2024-09-20 14:57:40 -07:00
Jeff McCune
16a4f89c2f doc: clarify story in the deploy a service guide pass 2
Focus on the migration team and platform team.
2024-09-20 11:12:41 -07:00
Jeff McCune
f15dea5ee7 doc: clarify story in the deploy a service guide
Focus on the migration team and platform team.
2024-09-20 11:02:41 -07:00
Jeff McCune
a3bbadd1f5 doc: apply manifests for deploy a service guide
This gets us through to the end with podinfo deployed.  Need to tell the
story of the migration team a bit better though, working with the
platform team to expose the service.
2024-09-20 09:42:00 -07:00
Jeff McCune
30cbb0d537 doc: add deploy a service guide
Covers wrapping a helm chart with a Holos Component.
2024-09-19 21:31:45 -07:00
Jeff McCune
6041fd4d76 website: fix broken links 2024-09-19 09:51:27 -07:00
Jeff McCune
fec1de0004 website: holistic platform manager social card 2024-09-19 09:49:56 -07:00
Jeff McCune
6ad24a6eec package: rename schema api to author api (#248)
Schema API is unclear, Author API is more clear the API is intended for
component authors.
2024-09-19 08:51:01 -07:00
Jeff McCune
57dedc6450 website: clean up placeholders 2024-09-19 08:30:31 -07:00
Jeff McCune
8d2a9dd659 quickstart: re-focus on core concepts 2024-09-18 17:24:44 -07:00
Jeff McCune
e3c53f5655 website: tweak features on landing page for clarity 2024-09-18 13:44:44 -07:00
Jeff McCune
3b833cdacd website: update landing page to focus on platform management
Instead of package management.
2024-09-18 13:41:36 -07:00
Jeff McCune
31d1086345 render: shorten the component rendered in the output
It's too long for documentation.  Shorten it for clarity.

Result:

```
❯ holos render platform ./platform
rendered bank-accounts-db for cluster workload in 160.7245ms
rendered bank-ledger-db for cluster workload in 162.465625ms
rendered bank-userservice for cluster workload in 166.150417ms
rendered bank-ledger-writer for cluster workload in 168.075459ms
rendered bank-balance-reader for cluster workload in 172.492292ms
rendered bank-backend-config for cluster workload in 198.117916ms
rendered bank-secrets for cluster workload in 223.200042ms
rendered gateway for cluster workload in 124.841917ms
rendered httproutes for cluster workload in 131.86625ms
rendered bank-contacts for cluster workload in 154.463792ms
rendered bank-transaction-history for cluster workload in 159.968208ms
rendered bank-frontend for cluster workload in 325.24425ms
rendered app-projects for cluster workload in 110.577916ms
rendered ztunnel for cluster workload in 137.502792ms
rendered cni for cluster workload in 209.993375ms
rendered cert-manager for cluster workload in 172.933834ms
rendered external-secrets for cluster workload in 135.759792ms
rendered local-ca for cluster workload in 98.026708ms
rendered istiod for cluster workload in 403.050833ms
rendered argocd for cluster workload in 294.663167ms
rendered gateway-api for cluster workload in 228.47875ms
rendered namespaces for cluster workload in 113.586916ms
rendered base for cluster workload in 533.76675ms
rendered external-secrets-crds for cluster workload in 529.053375ms
rendered crds for cluster workload in 931.180458ms
rendered platform in 1.248310167s
```

Previously:

```
❯ holos render platform ./platform
rendered projects/bank-of-holos/backend/components/bank-ledger-db for cluster workload in 158.534875ms
rendered projects/bank-of-holos/backend/components/bank-accounts-db for cluster workload in 159.836166ms
rendered projects/bank-of-holos/backend/components/bank-userservice for cluster workload in 160.360667ms
rendered projects/bank-of-holos/backend/components/bank-balance-reader for cluster workload in 169.478584ms
rendered projects/bank-of-holos/backend/components/bank-ledger-writer for cluster workload in 169.437833ms
rendered projects/bank-of-holos/backend/components/bank-backend-config for cluster workload in 182.089333ms
rendered projects/bank-of-holos/security/components/bank-secrets for cluster workload in 196.502792ms
rendered projects/platform/components/istio/gateway for cluster workload in 122.273083ms
rendered projects/bank-of-holos/frontend/components/bank-frontend for cluster workload in 307.573584ms
rendered projects/platform/components/httproutes for cluster workload in 149.631583ms
rendered projects/bank-of-holos/backend/components/bank-contacts for cluster workload in 153.529708ms
rendered projects/bank-of-holos/backend/components/bank-transaction-history for cluster workload in 165.375667ms
rendered projects/platform/components/app-projects for cluster workload in 107.253958ms
rendered projects/platform/components/istio/ztunnel for cluster workload in 137.22275ms
rendered projects/platform/components/istio/cni for cluster workload in 233.980958ms
rendered projects/platform/components/cert-manager for cluster workload in 171.966958ms
rendered projects/platform/components/external-secrets for cluster workload in 134.207792ms
rendered projects/platform/components/istio/istiod for cluster workload in 403.19ms
rendered projects/platform/components/local-ca for cluster workload in 97.544708ms
rendered projects/platform/components/argocd/argocd for cluster workload in 289.577208ms
rendered projects/platform/components/gateway-api for cluster workload in 218.290458ms
rendered projects/platform/components/namespaces for cluster workload in 109.534125ms
rendered projects/platform/components/istio/base for cluster workload in 526.32525ms
rendered projects/platform/components/external-secrets-crds for cluster workload in 523.7495ms
rendered projects/platform/components/argocd/crds for cluster workload in 1.002546375s
rendered platform in 1.312824333s
```
2024-09-17 11:40:14 -07:00
Jeff McCune
b737543c13 version: 0.95.0 2024-09-17 08:06:38 -07:00
Jeff McCune
8e150ee0d7 generate: fix external-secrets always out of sync in argocd
Without this patch ArgoCD treats the Application as constantly out of
sync.  This is also a good example of how to patch an arbitrary
component, though it patches the core BuildPlan itself now.  If this is
widely used, it would be nice to add this behavior to the schema api
(aka author api).
2024-09-16 21:22:23 -07:00
Jeff McCune
117a2a886d generate: fix istio-base always out of sync in argocd
Without this patch ArgoCD treats the Application as constantly out of
sync.  This is also a good example of how to patch an arbitrary
component, though it patches the core BuildPlan itself now.  If this is
widely used, it would be nice to add this behavior to the schema api
(aka author api).
2024-09-16 20:46:29 -07:00
Jeff McCune
79b41dfbf5 generate: fix istiod always out of sync in argocd
Without this patch ArgoCD treats the Application as constantly out of
sync.  This is also a good example of how to patch an arbitrary
component, though it patches the core BuildPlan itself now.  If this is
widely used, it would be nice to add this behavior to the schema api
(aka author api).
2024-09-16 20:38:35 -07:00
Jeff McCune
55562f9d83 generate: move istio-k3d schematic to projects structure
To match the current layout of the guide platform we're using for the
guides.
2024-09-16 20:02:23 -07:00
Jeff McCune
e0a636f183 generate: move gateway-api schematic to projects structure
To match the current layout of the guide platform we're using for the
guides.
2024-09-16 19:53:28 -07:00
Jeff McCune
1fa74214cf generate: move istio schematic to projects structure
To match the current layout of the guide platform we're using for the
guides.
2024-09-16 19:51:02 -07:00
Jeff McCune
e5851cac57 generate: fix bank of holos connection reset
Without this patch browsing https://bank.holos.localhost frequently gets
connection reset errors.  These errors are caused by the frontend
deployment redirecting the browser to http, which is not enabled on the
Gateway we use in the guides.

This patch sets the scheme to https which corrects the problems.

See https://github.com/GoogleCloudPlatform/bank-of-anthos/issues/478
2024-09-16 19:29:21 -07:00
Jeff McCune
4a26662b92 generate: add bank-contacts
Needed to load the home page of the Bank of Holos demo.
2024-09-16 19:29:02 -07:00
Jeff McCune
6bc6888ffc generate: add bank-transaction-history
Needed to load the home page of the Bank of Holos demo.
2024-09-16 17:00:15 -07:00
Jeff McCune
dab1f305e1 generate: add bank-balance-reader
Needed to load the home page of the Bank of Holos demo.
2024-09-16 16:52:52 -07:00
Jeff McCune
fbe79dd0af generate: add bank-ledger-db and bank-ledger-writer
Needed to load the home page of the Bank of Holos demo.
2024-09-16 16:46:35 -07:00
Jeff McCune
6d6829b149 generate: refactor bank backend config to a component
To fix ArgoCD SharedResourceWarning.
2024-09-16 16:18:31 -07:00
Jeff McCune
971a3fa280 generate: fix accounts-db using wrong service account
Needs to match the bank-of-holos service account name.
2024-09-16 15:58:50 -07:00
Jeff McCune
7632344cd1 generate: add bank-accounts-db needed by userservice
With this patch the frontend, accounts-db, and userservice all start and
become ready.

The user can log in, but on redirecting to home the site can't be
reached.
2024-09-16 15:51:29 -07:00
Jeff McCune
42067748ad generate: add bank userservice to backend 2024-09-16 15:49:25 -07:00
Jeff McCune
340c3484e5 generate: refactor how the bank jwt-key is created
This makes the example more re-usable, reader need only change the
SecretName, the bash script, and replace the #BankOfHolos references.
2024-09-16 15:02:12 -07:00
Jeff McCune
250238c286 generate: add secret store and external secret to bank-of-holos
Rather than commit the jwt private key to version control like upstream
does, we use a SecretStore and ExternalSecret to sync the secret
generated by the security team in the bank-security namespace.

With this patch the SecretStore validates and the ExternalSecret
automatically syncs the secret from the bank-security namespace to the
bank-frontend namespace.

```
❯ k get ss
NAME            AGE   STATUS   CAPABILITIES   READY
bank-security   1s    Valid    ReadWrite      True

❯ k get es
NAME      STORE           REFRESH INTERVAL   STATUS         READY
jwt-key   bank-security   5s                 SecretSynced   True
```

The pod start successfully.

```
❯ k get pods
NAME                        READY   STATUS    RESTARTS   AGE
frontend-646d797d6b-7jhrx   1/1     Running   0          2m39s

❯ k logs frontend-646d797d6b-7jhrx
{"timestamp": "2024-09-16 21:44:47", "message": "info | Starting gunicorn 22.0.0", "severity": "INFO"}
{"timestamp": "2024-09-16 21:44:47", "message": "info | Listening at: http://0.0.0.0:8080 (7)", "severity": "INFO"}
{"timestamp": "2024-09-16 21:44:47", "message": "info | Using worker: gthread", "severity": "INFO"}
{"timestamp": "2024-09-16 21:44:47", "message": "info | Booting worker with pid: 8", "severity": "INFO"}
{"timestamp": "2024-09-16 21:44:57", "message": "create_app | Unable to retrieve cluster name from metadata server metadata.google.internal.", "severity": "WARNING"}
{"timestamp": "2024-09-16 21:44:57", "message": "create_app | Unable to retrieve zone from metadata server metadata.google.internal.", "severity": "WARNING"}
{"timestamp": "2024-09-16 21:44:57", "message": "create_app | Starting frontend service.", "severity": "INFO"}
{"timestamp": "2024-09-16 21:44:57", "message": "create_app | 🚫 Tracing disabled.", "severity": "INFO"}
{"timestamp": "2024-09-16 21:44:57", "message": "create_app | Platform is set to 'local'", "severity": "INFO"}
```
2024-09-16 14:47:33 -07:00
Jeff McCune
a223e2b426 generate: fix duplicate external secrets crds
They're handled outside the helm chart to make upgrades easier.
2024-09-16 13:39:04 -07:00
Jeff McCune
63a7da02e7 generate: add external secrets operator
Need this to sync secrets into the bank-frontend and bank-backend
namespace from the bank-security namespace.
2024-09-16 13:29:03 -07:00
Jeff McCune
569f827e30 speed up argocd crds with raw urls 2024-09-16 13:29:02 -07:00
Jeff McCune
4a656db2ec render: log total render platform time 2024-09-16 13:29:02 -07:00
Jeff McCune
77b0933961 generate: add httproute for bank.holos.localhost
Expose Service frontend in the bank-frontend namespace via httproute
https://bank.holos.localhost

Organize into frontend, backend, security projects to align with three
teams who would each own this work.

remove secret from version control

Google added the secret to version control but we can generate the
secret in-cluster.  Holos makes it easier to manage the ExternalSecret
or RoleBinding necessary to get it in the right place.
2024-09-16 12:46:00 -07:00
Jeff McCune
3b796cfbbd generate: add bank-of-holos frontend
We need a way to demonstrate the value Holos offers in a platform team
managing projects for other teams.  This patch addresses the need by
establishing the bank-of-holos schematic, which is a port of the Bank of
Anthos project to Holos.

This patch adds only the frontend to get the process started.  As of
this patch the frontend pod starts and becomes ready but is not exposed
via HTTPRoute.

Refer to https://github.com/GoogleCloudPlatform/bank-of-anthos/
2024-09-15 22:08:28 -07:00
Jeff McCune
8a7a010b94 version 0.94.0 2024-09-15 15:41:17 -07:00
Jeff McCune
2454f6e9ee generate: app-projects to organize ArgoCD Applications into Projects
Previously all generated ArgoCD Application resources go into the
default project following the Quickstart guide.  The configuration code
is being organized into the concept of projects in the filesystem, so we
want to the GitOps configuration to also reflect this concept of
projects.

This patch extends the ArgoConfig user facing schema to accept a project
string.  The app-projects component automatically manages AppProject
resources in the argocd namespace for each of the defined projects.

This allows CUE configuration in the a project directory to specify the
project name so that all Applications are automatically assigned to the
correct project.
2024-09-15 14:57:13 -07:00
Jeff McCune
63d00bfddf schema: handle ArgoConfig for all component kinds
Providing ArgoConfig only works with the Helm kind without this patch.
This is a problem because we want to produce an Application for every
supported component kind when rendering the platform.

This patch threads the ArgoConfig struct described in the Quickstart
guide through every supported component kind.
2024-09-15 13:09:07 -07:00
Jeff McCune
f34da6c24e generate: add schematic for manage a project guide gitops
This patch configures the platform to generate Application resources for
all of the components in the manage a project guide.
2024-09-15 12:57:51 -07:00
Jeff McCune
1d98069b73 generate: add httproutes schematic
Define a place for components to register HTTPRoute resources the
platform team needs to manage in the Gateway namespace.

The files are organized to delegate to the platform team.

This patch also fixes the naming of the argocd component so that the
Service is argocd-server instead of argo-cd-argocd-server.
2024-09-15 12:32:45 -07:00
Jeff McCune
e956b64d04 schematic: comment how kustomization of argocd crds works
This is the only example we have right now of producing a kustomization
entirely from CUE.
2024-09-15 09:37:42 -07:00
Jeff McCune
054d33b498 builder: add kustomization post-processing to kubernetes build plans (#246)
Holos does not post-process a KubernetesObjects core package build plan
with kustomize.  This is necessary to pass the ArgoCD version through to
Kustomize to fetch the correct crds.

This patch enables the kustomization and provides an example in the
argocd schematic.

Result: The KubernetesObjects component doesn't actually have any
resources defined, so holos creates an empty `build-plan-resources.yaml`
file.  This is fine, the kustomize post-processing adds the actual
resources via https URL passing in the correct ArgoCD version.

```
❯ holos render component --cluster-name=workload ./projects/platform/components/argocd/crds --log-level=debug --log-format=text
9:20AM DBG config.go:166 finalized config from flags version=0.93.4 state=finalized
9:20AM DBG builder.go:234 cue: building instances version=0.93.4
9:20AM DBG builder.go:251 cue: validating instance version=0.93.4 dir=/Users/jeff/Holos/holos-manage-a-project-guide/projects/platform/components/argocd/crds
9:20AM DBG builder.go:256 cue: decoding holos build plan version=0.93.4 dir=/Users/jeff/Holos/holos-manage-a-project-guide/projects/platform/components/argocd/crds
9:20AM DBG builder.go:270 cue: discriminated build kind: BuildPlan version=0.93.4 dir=/Users/jeff/Holos/holos-manage-a-project-guide/projects/platform/components/argocd/crds kind=BuildPlan apiVersion=v1alpha3
9:20AM DBG builder.go:314 allocated results slice version=0.93.4 cap=1
9:20AM DBG result.go:156 wrote: /var/folders/22/zt67pphj6h1fgknqfy23ppl80000gn/T/holos.kustomize3526125146/build-plan-resources.yaml version=0.93.4 op=write path=/var/folders/22/zt67pphj6h1fgknqfy23ppl80000gn/T/holos.kustomize3526125146/build-plan-resources.yaml bytes=0
9:20AM DBG result.go:169 wrote: /var/folders/22/zt67pphj6h1fgknqfy23ppl80000gn/T/holos.kustomize3526125146/kustomization.yaml version=0.93.4 op=write path=/var/folders/22/zt67pphj6h1fgknqfy23ppl80000gn/T/holos.kustomize3526125146/kustomization.yaml bytes=174
9:20AM DBG run.go:40 running: kubectl version=0.93.4 name=kubectl args="[kustomize /var/folders/22/zt67pphj6h1fgknqfy23ppl80000gn/T/holos.kustomize3526125146]"
9:20AM DBG remove.go:16 tmp: removed version=0.93.4 path=/var/folders/22/zt67pphj6h1fgknqfy23ppl80000gn/T/holos.kustomize3526125146
9:20AM DBG builder.go:350 returning results version=0.93.4 len=1
9:20AM DBG result.go:214 out: wrote deploy/clusters/workload/components/argocd-crds/argocd-crds.gen.yaml version=0.93.4 action=write path=deploy/clusters/workload/components/argocd-crds/argocd-crds.gen.yaml status=ok
9:20AM INF render.go:79 rendered argocd-crds version=0.93.4 cluster=workload name=argocd-crds status=ok action=rendered
```

Closes: #246
2024-09-15 09:25:25 -07:00
Jeff McCune
f2f75a4e00 generate: fix argo-cd component 2024-09-15 09:06:53 -07:00
Jeff McCune
a0cf73faf9 generate: remove argocd kustomization.yaml 2024-09-15 09:00:41 -07:00
Jeff McCune
d74655c632 generate: add argocd schematic
Using the projects layout.

This patch also includes a method to pass a version to a Kustomization.
2024-09-15 08:57:57 -07:00
Jeff McCune
b8019429b8 docs: add manage a project guide draft (#242)
Initial draft of the Manage a Project guide focused on how a development
team can self serve resources provided by a platform team.
2024-09-14 15:35:12 -07:00
Jeff McCune
9c08214118 docs: add the outline of the projects guide (#242) 2024-09-14 13:43:31 -07:00
Jeff McCune
f58d791e03 api: move #Resources to package holos
Previously, the #Resources struct listing valid resources to use with
APIObjects in each of the components types was closed.  This made it
very difficult for users to mix in new resources and use the Kubernetes
component kind.

This patch moves the definition of the valid resources to package holos
from the schema API.  The schema still enforces some light constraints,
but doesn't keep the struct closed.

A new convention is introduced in the form of configuring all components
using _ComponentConfig defined at the root, then unifying this struct
with all of the component kinds.  See schema.gen.cue for how this works.

This approach enables mixing in ArgoCD applications to all component
kinds, not just Helm as was done previously.  Similarly, the
user-constrained #Resources definition unifies with all component kinds.

It's OK to leave the yaml.Marshall in the schema API.  The user
shouldn't ever have to deal with #APIObjects, instead they should pass
Resources through the schema API which will use APIObjects to create
apiObjectMap for each component type and the BuildPlan.

This is still more awkward than I want, but it's a good step in the
right direction.
2024-09-13 16:43:12 -07:00
Jeff McCune
836033e16a docs: move istio-gateway to a separate schematic
Without this patch the istio-gateway component isn't functional, the
HTTPRoute created for httpbin isn't programmed correctly.  There is no
Gateway resource, just a deployment created by the istio helm chart.

This patch replaces the helm chart with a Gateway resource as was done
previously in the k3d platform schematic.

This patch also simplifies the certificate management to issue a single
cert valid for the platform domain and a wildcard.  We intentionally
avoid building a dynamic Gateway.spec.listeners structure to keep the
expose a service guide relatively simple and focused on getting started
with Holos.
2024-09-13 11:13:18 -07:00
Jeff McCune
77279d9baf docs: add httpbin routes section to expose a service guide
This patch adds the httpbin routes component.  It's missing the
Certificate component, the next step is to wire up automatic certificate
management in the gateway configuration, which is a prime use case for
holos.  Similar to how we register components and namespaces, we'll
register certificates.

This patch also adds the #Platform.Domain field to the user facing
schema API.  We previously stored the domain in the Model but it makes
sense to lift it up to the Platform and have a sensible default value
for it.

Another example of #237 needing to be addressed soon.
2024-09-12 17:35:06 -07:00
Jeff McCune
bf19aee1a7 docs: add httpbin workload section to expose a service
This patch manages the httpbin Deployment, Service, and ReferenceGrant.
The remaining final step is to expose the service with an HTTPRoute and
Certificate.

We again needed to add a field to the schema APIObjects to get this to
work.  We need to fix #237 soon.  We'll need to do it again for the
HTTPRoute and Certificate resources.
2024-09-12 16:55:09 -07:00
Jeff McCune
4de88b3155 docs: insert cert-manager after namespaces in expose a service
The progression of namespaces, cert-manager, then gateway api and istio
makes much more sense than the previous progression of gateway api,
namespaces, istio.

cert-manager builds nicely on top of namespaces.  gateway api are only
crds necessary for istio.

This patch also adds the local-ca component which surfaces issue #237
The Kubernetes APIObjects are unnecessarily constrained to resources we
define in the schema.  We need to move the marshal code into package
holos so the user can add their own resource kinds.
2024-09-12 15:20:08 -07:00
Jeff McCune
6f39cc6fdc docs: add istio section to expose-a-service
This patch adds Istio to the Expose a Service documentation and
introduces new concepts.  The Kubernetes build plan schema, the
namespaces component, and an example of how to safely re-use Helm values
from the root to multiple leaf components.

fix: istio cni not ready on k3d
---

The istio-k3d component embedded into holos fixes the cni pod not
becoming ready with our k3d local cluster guide.  The pod log error this
fixes is:

    configuration requires updates, (re)writing CNI config file at "": no networks found in /host/etc/cni/net.d
    Istio CNI is configured as chained plugin, but cannot find existing CNI network config: no networks found in /host/etc/cni/net.d
    Waiting for CNI network config file to be written in /host/etc/cni/net.d...

[Platform k3d]: https://istio.io/latest/docs/ambient/install/platform-prerequisites/#k3d

docs: clarify how to reset the local cluster
---

This is something we do all the time while developing and documenting,
so make it easy and fast to reset the cluster to a known good state.
2024-09-12 10:36:56 -07:00
Jeff McCune
e410563f82 docs: add namespaces to expose a service guide
This patch adds the schema api for the Kubernetes build plan, which
produces plain API resources directly from CUE.  It's needed for the
namespaces component which is foundational to many of our guides.

The first guide that needs this is the expose a service guide, we need
to register the namespaces from the istio component.
2024-09-11 17:22:01 -07:00
Jeff McCune
0a53bef72a docs: apply the gateway-api in the expose a service doc
This patch completes the first draft of the Gateway API section.
2024-09-11 14:31:02 -07:00
Jeff McCune
02a450e597 api: clarify Name field of Helm and Kustomize schema 2024-09-11 14:09:13 -07:00
Jeff McCune
e1222cf367 docs: add the gateway-api to the expose-a-service doc
The Expose a Service doc is meant to be the second step after the
Quickstart doc.  This commit adds the section describing how to install
the Gateway API.

The Kustomize build plan is introduced at this point in a similar way
the Helm build plan was introduced in the quickstart.
2024-09-11 14:03:40 -07:00
Jeff McCune
740a3d21a1 generate: add schematic for a workload-cluster
We need an easy way to help people add a workload cluster to their
workload fleet when working through the guides.  Generated platforms
should not define any clusters so they can be reused with multiple
guides.

This patch adds a simple component schematic that drops a root cue file
to define a workload cluster named workload.

The result is the following sequence renders the Gateway API when run
from an empty directory.

    holos generate platform guide
    holos generate component workload-cluster
    holos generate component gateway-api
    holos render platform ./platform

Without this patch nothing is rendered because there are no workload
clusters in the base guide platform.
2024-09-11 13:23:36 -07:00
Jeff McCune
1114b65a47 schema: remove management cluster from standard fleet
Having the management cluster hard coded into the definition of the
standard fleets is problematic for guides that don't need a management
cluster.

Define the fleets, but leave the set of clusters empty until they're
needed.
2024-09-11 13:12:44 -07:00
239 changed files with 30061 additions and 1346 deletions

View File

@@ -5,48 +5,93 @@
"mdx"
],
"words": [
"acraccesstokens",
"admissionregistration",
"anthos",
"apiextensions",
"applicationset",
"applicationsets",
"appproject",
"appprojects",
"argoproj",
"authcode",
"authorizationpolicies",
"authpolicy",
"authproxy",
"authroutes",
"balancereader",
"buildplan",
"cainjector",
"CAROOT",
"certificaterequests",
"certificatesigningrequests",
"clsx",
"clusterexternalsecrets",
"clusterissuer",
"clusterissuers",
"clusterrole",
"clusterrolebinding",
"clustersecretstores",
"CNCF",
"CODEOWNERS",
"configmap",
"cookiesecret",
"coredns",
"corev",
"CRD's",
"crds",
"creds",
"crossplane",
"cuecontext",
"cuelang",
"customresourcedefinition",
"daemonset",
"destinationrules",
"devicecode",
"dnsmasq",
"dscacheutil",
"ecrauthorizationtokens",
"entgo",
"envoyfilters",
"errgroup",
"etcdsnapshotfiles",
"externalsecret",
"externalsecrets",
"fctr",
"fieldmaskpb",
"flushcache",
"gatewayclasses",
"gcraccesstokens",
"gendoc",
"ggnpl",
"ghaction",
"githubaccesstokens",
"gitops",
"godoc",
"golangci",
"goreleaser",
"grpcreflect",
"grpcroutes",
"grpcurl",
"healthz",
"helmchartconfigs",
"helmcharts",
"Hiera",
"holos",
"holoslogger",
"horizontalpodautoscaler",
"Hostnames",
"httpbin",
"httproute",
"httproutes",
"Infima",
"isatty",
"istiod",
"jbrx",
"jeffmccune",
"jetstack",
"Jsonnet",
"kfbh",
"killall",
"kubeadm",
"kubeconfig",
@@ -55,39 +100,70 @@
"Kustomizations",
"kustomize",
"ldflags",
"leaderelection",
"ledgerwriter",
"libnss",
"loadbalancer",
"mattn",
"mccutchen",
"mindmap",
"mktemp",
"msqbn",
"mtls",
"Multicluster",
"mutatingwebhookconfiguration",
"mxcl",
"myhostname",
"nameserver",
"nolint",
"organizationconnect",
"orgid",
"otelconnect",
"Parentspanid",
"pcjc",
"peerauthentications",
"pflag",
"pipefail",
"PKCE",
"platformconnect",
"podcli",
"poddisruptionbudget",
"podinfo",
"portmapping",
"promhttp",
"protobuf",
"protojson",
"proxyconfigs",
"Pulumi",
"pushsecrets",
"putenv",
"qjbp",
"quickstart",
"readyz",
"referencegrant",
"referencegrants",
"requestauthentications",
"retryable",
"rolebinding",
"ropc",
"seccomp",
"SECRETKEY",
"secretstore",
"secretstores",
"serverlb",
"serverside",
"serviceaccount",
"servicebindings",
"serviceentries",
"spanid",
"spiffe",
"startupapicheck",
"statefulset",
"stefanprodan",
"struct",
"structpb",
"subjectaccessreviews",
"svclb",
"systemconnect",
"tablewriter",
"Tiltfile",
@@ -98,14 +174,24 @@
"Tokener",
"Traceid",
"traefik",
"transactionhistory",
"uibutton",
"unstage",
"untar",
"Upsert",
"urandom",
"usecases",
"userconnect",
"userdata",
"userservice",
"validatingwebhookconfiguration",
"vaultdynamicsecrets",
"virtualservices",
"wasmplugins",
"workloadentries",
"workloadgroups",
"zerolog",
"zitadel"
"zitadel",
"ztunnel"
]
}

View File

@@ -11,18 +11,29 @@ import (
//go:generate ../../../hack/gendoc
// Component represents the fields common the different kinds of component. All
// components have a name, support mixing in resources, and produce a BuildPlan.
type ComponentFields struct {
// Name represents the Component name.
Name string
// Resources are kubernetes api objects to mix into the output.
Resources map[string]any
// ArgoConfig represents the ArgoCD GitOps configuration for this Component.
ArgoConfig ArgoConfig
// BuildPlan represents the derived BuildPlan for the Holos cli to render.
BuildPlan core.BuildPlan
}
// Helm provides a BuildPlan via the Output field which contains one HelmChart
// from package core. Useful as a convenience wrapper to render a HelmChart
// with optional mix-in resources and Kustomization post-processing.
type Helm struct {
// Name represents the chart name.
Name string
ComponentFields `json:",inline"`
// Version represents the chart version.
Version string
// Namespace represents the helm namespace option when rendering the chart.
Namespace string
// Resources are kubernetes api objects to mix into the output.
Resources map[string]any `cue:"{...}"`
// Repo represents the chart repository
Repo struct {
@@ -57,27 +68,23 @@ type Helm struct {
// KustomizeResources represents additional resources files to include in the
// kustomize resources list.
KustomizeResources map[string]any `cue:"{[string]: {...}}"`
// ArgoConfig represents the ArgoCD GitOps configuration for this Component.
ArgoConfig ArgoConfig
// Output represents the derived BuildPlan for the Holos cli to render.
Output core.BuildPlan
}
// Resources represents the default schema for a Kubernetes API object resource.
// For example, a Service, Namespace or Deployment. The top level key is the
// kind of resource so default behavior and strict schema enforcement may be
// enforced for the kind. The second level keys are an arbitrary internal
// label, which serves as the default value for the resource metadata name
// field, but may differ for situations where the same resource kind and name
// are managed in different namespaces.
//
// Refer to [definitions.cue] for the CUE schema definition as an example to
// build on when defining your own Components.
//
// [definitions.cue]: https://github.com/holos-run/holos/blob/main/internal/generate/platforms/cue.mod/pkg/github.com/holos-run/holos/api/schema/v1alpha3/definitions.cue#L9
// type Resources map[string]map[string]any
// Kustomize provides a BuildPlan via the Output field which contains one
// KustomizeBuild from package core.
type Kustomize struct {
ComponentFields `json:",inline"`
// Kustomization represents the kustomize build plan for holos to render.
Kustomization core.KustomizeBuild
}
// Kubernetes provides a BuildPlan via the Output field which contains inline
// API Objects provided directly from CUE.
type Kubernetes struct {
ComponentFields `json:",inline"`
// Objects represents the kubernetes api objects for the Component.
Objects core.KubernetesObjects
}
// ArgoConfig represents the ArgoCD GitOps configuration for a Component.
// Useful to define once at the root of the Platform configuration and reuse
@@ -99,6 +106,8 @@ type ArgoConfig struct {
// Application.spec.source.targetRevision field. Defaults to the branch named
// main.
TargetRevision string `cue:"string | *\"main\""`
// AppProject represents the ArgoCD Project to associate the Application with.
AppProject string `cue:"string | *\"default\""`
}
// Cluster represents a cluster managed by the Platform.
@@ -129,7 +138,7 @@ type StandardFleets struct {
// Workload represents a Fleet of zero or more workload Clusters.
Workload Fleet `json:"workload" cue:"{name: \"workload\"}"`
// Management represents a Fleet with one Cluster named management.
Management Fleet `json:"management" cue:"{name: \"management\", clusters: management: _}"`
Management Fleet `json:"management" cue:"{name: \"management\"}"`
}
// Platform is a convenience structure to produce a core Platform specification
@@ -147,4 +156,71 @@ type Platform struct {
// Output represents the core Platform spec for the holos cli to iterate over
// and render each listed Component, injecting the Model.
Output core.Platform
// Domain represents the primary domain the Platform operates in. This field
// is intended as a sensible default for component authors to reference and
// platform operators to define.
Domain string `cue:"string | *\"holos.localhost\""`
}
// Organization represents organizational metadata useful across the platform.
type Organization struct {
Name string
DisplayName string
Domain string
}
// OrganizationStrict represents organizational metadata useful across the
// platform. This is an example of using CUE regular expressions to constrain
// and validate configuration.
type OrganizationStrict struct {
Organization `json:",inline"`
// Name represents the organization name as a resource name. Must be 63
// characters or less. Must start with a letter. May contain non-repeating
// hyphens, letters, and numbers. Must end with a letter or number.
Name string `cue:"=~ \"^[a-z][0-9a-z-]{1,61}[0-9a-z]$\" & !~ \"--\""`
// DisplayName represents the human readable organization name.
DisplayName string `cue:"=~ \"^[0-9A-Za-z][0-9A-Za-z ]{2,61}[0-9A-Za-z]$\" & !~ \" \""`
}
// Projects represents projects managed by the platform team for use by other
// teams using the platform.
type Projects map[core.NameLabel]Project
// Project represents logical grouping of components owned by one or more teams.
// Useful for the platform team to manage resources for project teams to use.
type Project struct {
// Name represents project name.
Name string
// Owner represents the team who own this project.
Owner Owner
// Namespaces represents the namespaces assigned to this project.
Namespaces map[core.NameLabel]Namespace
// Hostnames represents the host names to expose for this project.
Hostnames map[core.NameLabel]Hostname
}
// Owner represents the owner of a resource. For example, the name and email
// address of an engineering team.
type Owner struct {
Name string
Email string
}
// Namespace represents a Kubernetes namespace.
type Namespace struct {
Name string
}
// Hostname represents the left most dns label of a domain name.
type Hostname struct {
// Name represents the subdomain to expose, e.g. "www"
Name string
// Namespace represents the namespace metadata.name field of backend object
// reference.
Namespace string
// Service represents the Service metadata.name field of backend object
// reference.
Service string
// Port represents the Service port of the backend object reference.
Port int
}

View File

@@ -14,6 +14,12 @@ import "google.golang.org/protobuf/types/known/structpb"
// output.
type InternalLabel string
// NameLabel is a unique identifier useful to convert a CUE struct to a list
// when the values have a Name field with a default value. This type is
// intended to indicate the common use case of converting a struct to a list
// where the Name field of the value aligns with the struct field name.
type NameLabel string
// Kind is a kubernetes api object kind. Defined as a type for clarity and type
// checking.
type Kind string

265
api/core/v1alpha4/types.go Normal file
View File

@@ -0,0 +1,265 @@
// Package v1alpha4 contains the core API contract between the holos cli and CUE
// configuration code. Platform designers, operators, and software developers
// use this API to write configuration in CUE which `holos` loads. The overall
// shape of the API defines imperative actions `holos` should carry out to
// render the complete yaml that represents a Platform.
//
// [Platform] defines the complete configuration of a platform. With the holos
// reference platform this takes the shape of one management cluster and at
// least two workload clusters.
//
// Each holos component path, e.g. `components/namespaces` produces exactly one
// [BuildPlan] which produces an [Artifact] collection. An [Artifact] is a
// fully rendered manifest produced from a [Transformer] sequence, which
// transforms a [Generator] collection.
package v1alpha4
//go:generate ../../../hack/gendoc
// BuildPlan represents a build plan for holos to execute. Each [Platform]
// component produces exactly one BuildPlan.
//
// One or more [Artifact] files are produced by a BuildPlan, representing the
// fully rendered manifests for the Kubernetes API Server.
type BuildPlan struct {
// Kind represents the type of the resource.
Kind string `json:"kind" cue:"\"BuildPlan\""`
// APIVersion represents the versioned schema of the resource.
APIVersion string `json:"apiVersion" cue:"string | *\"v1alpha4\""`
// Metadata represents data about the resource such as the Name.
Metadata Metadata `json:"metadata"`
// Spec specifies the desired state of the resource.
Spec BuildPlanSpec `json:"spec"`
}
// BuildPlanSpec represents the specification of the [BuildPlan].
type BuildPlanSpec struct {
// Component represents the component that produced the build plan.
// Represented as a path relative to the platform root.
Component string `json:"component"`
// Disabled causes the holos cli to disregard the build plan.
Disabled bool `json:"disabled,omitempty"`
// Artifacts represents the artifacts for holos to build.
Artifacts []Artifact `json:"artifacts"`
}
// Artifact represents one fully rendered manifest produced by a [Transformer]
// sequence, which transforms a [Generator] collection. A [BuildPlan] produces
// an [Artifact] collection.
//
// Each Artifact produces one manifest file artifact. Generator Output values
// are used as Transformer Inputs. The Output field of the final [Transformer]
// should have the same value as the Artifact field.
//
// When there is more than one [Generator] there must be at least one
// [Transformer] to combine outputs into one Artifact. If there is a single
// Generator, it may directly produce the Artifact output.
//
// An Artifact is processed concurrently with other artifacts in the same
// [BuildPlan]. An Artifact should not use an output from another Artifact as
// an input. Each [Generator] may also run concurrently. Each [Transformer] is
// executed sequentially starting after all generators have completed.
//
// Output fields are write-once. It is an error for multiple Generators or
// Transformers to produce the same Output value within the context of a
// [BuildPlan].
type Artifact struct {
Artifact FilePath `json:"artifact,omitempty"`
Generators []Generator `json:"generators,omitempty"`
Transformers []Transformer `json:"transformers,omitempty"`
Skip bool `json:"skip,omitempty"`
}
// Generator generates an intermediate manifest for a [Artifact].
//
// Each Generator in a [Artifact] must have a distinct Output value for a
// [Transformer] to reference.
//
// Refer to [Resources], [Helm], and [File].
type Generator struct {
// Kind represents the kind of generator. Must be Resources, Helm, or File.
Kind string `json:"kind" cue:"\"Resources\" | \"Helm\" | \"File\""`
// Output represents a file for a Transformer or Artifact to consume.
Output FilePath `json:"output"`
// Resources generator. Ignored unless kind is Resources. Resources are
// stored as a two level struct. The top level key is the Kind of resource,
// e.g. Namespace or Deployment. The second level key is an arbitrary
// InternalLabel. The third level is a map[string]any representing the
// Resource.
Resources Resources `json:"resources,omitempty"`
// Helm generator. Ignored unless kind is Helm.
Helm Helm `json:"helm,omitempty"`
// File generator. Ignored unless kind is File.
File File `json:"file,omitempty"`
}
// Resource represents one kubernetes api object.
type Resource map[string]any
// Resources represents a kubernetes resources [Generator] from CUE.
type Resources map[Kind]map[InternalLabel]Resource
// File represents a simple single file copy [Generator]. Useful with a
// [Kustomize] [Transformer] to process plain manifest files stored in the
// component directory. Multiple File generators may be used to transform
// multiple resources.
type File struct {
// Source represents a file sub-path relative to the component path.
Source FilePath `json:"source"`
}
// Helm represents a [Chart] manifest [Generator].
type Helm struct {
// Chart represents a helm chart to manage.
Chart Chart `json:"chart"`
// Values represents values for holos to marshal into values.yaml when
// rendering the chart.
Values Values `json:"values"`
// EnableHooks enables helm hooks when executing the `helm template` command.
EnableHooks bool `json:"enableHooks,omitempty"`
// Namespace represents the helm namespace flag
Namespace string `json:"namespace,omitempty"`
}
// Values represents [Helm] Chart values generated from CUE.
type Values map[string]any
// Chart represents a [Helm] Chart.
type Chart struct {
// Name represents the chart name.
Name string `json:"name"`
// Version represents the chart version.
Version string `json:"version"`
// Release represents the chart release when executing helm template.
Release string `json:"release"`
// Repository represents the repository to fetch the chart from.
Repository Repository `json:"repository,omitempty"`
}
// Repository represents a [Helm] [Chart] repository.
type Repository struct {
Name string `json:"name"`
URL string `json:"url"`
}
// Transformer transforms [Generator] manifests within a [Artifact].
type Transformer struct {
// Kind represents the kind of transformer. Must be Kustomize, or Join.
Kind string `json:"kind" cue:"\"Kustomize\" | \"Join\""`
// Inputs represents the files to transform. The Output of prior Generators
// and Transformers.
Inputs []FilePath `json:"inputs"`
// Output represents a file for a subsequent Transformer or Artifact to
// consume.
Output FilePath `json:"output"`
// Kustomize transformer. Ignored unless kind is Kustomize.
Kustomize Kustomize `json:"kustomize,omitempty"`
// Join transformer. Ignored unless kind is Join.
Join Join `json:"join,omitempty"`
}
// Join represents a [Join](https://pkg.go.dev/strings#Join) [Transformer].
// Useful for the common case of combining the output of [Helm] and [Resources]
// [Generator] into one [Artifact] when [Kustomize] is otherwise unnecessary.
type Join struct {
Separator string `json:"separator" cue:"string | *\"---\\n\""`
}
// Kustomize represents a kustomization [Transformer].
type Kustomize struct {
// Kustomization represents the decoded kustomization.yaml file
Kustomization Kustomization `json:"kustomization"`
// Files holds file contents for kustomize, e.g. patch files.
Files FileContentMap `json:"files,omitempty"`
}
// Kustomization represents a kustomization.yaml file for use with the
// [Kustomize] [Transformer]. Untyped to avoid tightly coupling holos to
// kubectl versions which was a problem for the Flux maintainers. Type checking
// is expected to happen in CUE against the kubectl version the user prefers.
type Kustomization map[string]any
// FileContent represents file contents.
type FileContent string
// FileContentMap represents a mapping of file paths to file contents.
type FileContentMap map[FilePath]FileContent
// FilePath represents a file path.
type FilePath string
// InternalLabel is an arbitrary unique identifier internal to holos itself.
// The holos cli is expected to never write a InternalLabel value to rendered
// output files, therefore use a InternalLabel when the identifier must be
// unique and internal. Defined as a type for clarity and type checking.
type InternalLabel string
// Kind is a discriminator. Defined as a type for clarity and type checking.
type Kind string
// NameLabel is a unique identifier useful to convert a CUE struct to a list
// when the values have a Name field with a default value. NameLabel indicates
// the common use case of converting a struct to a list where the Name field of
// the value aligns with the outer struct field name.
//
// For example:
//
// Outer: [NAME=_]: Name: NAME
type NameLabel string
// Platform represents a platform to manage. A Platform resource informs holos
// which components to build. The platform resource also acts as a container
// for the platform model form values provided by the PlatformService. The
// primary use case is to collect the cluster names, cluster types, platform
// model, and holos components to build into one resource.
type Platform struct {
// Kind is a string value representing the resource.
Kind string `json:"kind" cue:"\"Platform\""`
// APIVersion represents the versioned schema of this resource.
APIVersion string `json:"apiVersion" cue:"string | *\"v1alpha4\""`
// Metadata represents data about the resource such as the Name.
Metadata Metadata `json:"metadata"`
// Spec represents the specification.
Spec PlatformSpec `json:"spec"`
}
// Metadata represents data about the resource such as the Name.
type Metadata struct {
// Name represents the resource name.
Name string `json:"name"`
}
// PlatformSpec represents the specification of a [Platform]. Think of a
// platform spec as a [Component] collection for multiple kubernetes clusters
// combined with the user-specified Platform Model.
type PlatformSpec struct {
// Components represents a list of holos components to manage.
Components []Component `json:"components"`
}
// Component represents the complete context necessary to produce a [BuildPlan]
// from a [Platform] component.
//
// All of these fields are passed to the holos render component command using
// flags, which in turn are injected to CUE using tags. Field names should be
// used consistently through the platform rendering process for readability.
type Component struct {
// Name represents the name of the component, injected as a tag to set the
// BuildPlan metadata.name field. Necessary for clear user feedback during
// platform rendering.
Name string `json:"name"`
// Component represents the path of the component relative to the platform root.
Component string `json:"component"`
// Cluster is the cluster name to provide when rendering the component.
Cluster string `json:"cluster"`
// Environment for example, dev, test, stage, prod
Environment string `json:"environment,omitempty"`
// Model represents the platform model holos gets from from the
// PlatformService.GetPlatform rpc method and provides to CUE using a tag.
Model map[string]any `json:"model"`
// Tags represents cue tags to inject when rendering the component. The json
// struct tag names of other fields in this struct are reserved tag names not
// to be used in the tags collection.
Tags []string `json:"tags,omitempty"`
}

View File

@@ -1,5 +1,5 @@
import DocCardList from '@theme/DocCardList';
# Schema API
# Author API
<DocCardList />

View File

@@ -3,7 +3,7 @@
# v1alpha3
```go
import "github.com/holos-run/holos/api/schema/v1alpha3"
import "github.com/holos-run/holos/api/author/v1alpha3"
```
Package v1alpha3 contains CUE definitions intended as convenience wrappers around the core data types defined in package core. The purpose of these wrappers is to make life easier for platform engineers by reducing boiler plate code and generating component build plans in a consistent manner.
@@ -12,9 +12,19 @@ Package v1alpha3 contains CUE definitions intended as convenience wrappers aroun
- [type ArgoConfig](<#ArgoConfig>)
- [type Cluster](<#Cluster>)
- [type ComponentFields](<#ComponentFields>)
- [type Fleet](<#Fleet>)
- [type Helm](<#Helm>)
- [type Hostname](<#Hostname>)
- [type Kubernetes](<#Kubernetes>)
- [type Kustomize](<#Kustomize>)
- [type Namespace](<#Namespace>)
- [type Organization](<#Organization>)
- [type OrganizationStrict](<#OrganizationStrict>)
- [type Owner](<#Owner>)
- [type Platform](<#Platform>)
- [type Project](<#Project>)
- [type Projects](<#Projects>)
- [type StandardFleets](<#StandardFleets>)
@@ -41,6 +51,8 @@ type ArgoConfig struct {
// Application.spec.source.targetRevision field. Defaults to the branch named
// main.
TargetRevision string `cue:"string | *\"main\""`
// AppProject represents the ArgoCD Project to associate the Application with.
AppProject string `cue:"string | *\"default\""`
}
```
@@ -60,6 +72,24 @@ type Cluster struct {
}
```
<a name="ComponentFields"></a>
## type ComponentFields {#ComponentFields}
Component represents the fields common the different kinds of component. All components have a name, support mixing in resources, and produce a BuildPlan.
```go
type ComponentFields struct {
// Name represents the Component name.
Name string
// Resources are kubernetes api objects to mix into the output.
Resources map[string]any
// ArgoConfig represents the ArgoCD GitOps configuration for this Component.
ArgoConfig ArgoConfig
// BuildPlan represents the derived BuildPlan for the Holos cli to render.
BuildPlan core.BuildPlan
}
```
<a name="Fleet"></a>
## type Fleet {#Fleet}
@@ -80,14 +110,12 @@ Helm provides a BuildPlan via the Output field which contains one HelmChart from
```go
type Helm struct {
// Name represents the chart name.
Name string
ComponentFields `json:",inline"`
// Version represents the chart version.
Version string
// Namespace represents the helm namespace option when rendering the chart.
Namespace string
// Resources are kubernetes api objects to mix into the output.
Resources map[string]any `cue:"{...}"`
// Repo represents the chart repository
Repo struct {
@@ -122,12 +150,105 @@ type Helm struct {
// KustomizeResources represents additional resources files to include in the
// kustomize resources list.
KustomizeResources map[string]any `cue:"{[string]: {...}}"`
}
```
// ArgoConfig represents the ArgoCD GitOps configuration for this Component.
ArgoConfig ArgoConfig
<a name="Hostname"></a>
## type Hostname {#Hostname}
// Output represents the derived BuildPlan for the Holos cli to render.
Output core.BuildPlan
Hostname represents the left most dns label of a domain name.
```go
type Hostname struct {
// Name represents the subdomain to expose, e.g. "www"
Name string
// Namespace represents the namespace metadata.name field of backend object
// reference.
Namespace string
// Service represents the Service metadata.name field of backend object
// reference.
Service string
// Port represents the Service port of the backend object reference.
Port int
}
```
<a name="Kubernetes"></a>
## type Kubernetes {#Kubernetes}
Kubernetes provides a BuildPlan via the Output field which contains inline API Objects provided directly from CUE.
```go
type Kubernetes struct {
ComponentFields `json:",inline"`
// Objects represents the kubernetes api objects for the Component.
Objects core.KubernetesObjects
}
```
<a name="Kustomize"></a>
## type Kustomize {#Kustomize}
Kustomize provides a BuildPlan via the Output field which contains one KustomizeBuild from package core.
```go
type Kustomize struct {
ComponentFields `json:",inline"`
// Kustomization represents the kustomize build plan for holos to render.
Kustomization core.KustomizeBuild
}
```
<a name="Namespace"></a>
## type Namespace {#Namespace}
Namespace represents a Kubernetes namespace.
```go
type Namespace struct {
Name string
}
```
<a name="Organization"></a>
## type Organization {#Organization}
Organization represents organizational metadata useful across the platform.
```go
type Organization struct {
Name string
DisplayName string
Domain string
}
```
<a name="OrganizationStrict"></a>
## type OrganizationStrict {#OrganizationStrict}
OrganizationStrict represents organizational metadata useful across the platform. This is an example of using CUE regular expressions to constrain and validate configuration.
```go
type OrganizationStrict struct {
Organization `json:",inline"`
// Name represents the organization name as a resource name. Must be 63
// characters or less. Must start with a letter. May contain non-repeating
// hyphens, letters, and numbers. Must end with a letter or number.
Name string `cue:"=~ \"^[a-z][0-9a-z-]{1,61}[0-9a-z]$\" & !~ \"--\""`
// DisplayName represents the human readable organization name.
DisplayName string `cue:"=~ \"^[0-9A-Za-z][0-9A-Za-z ]{2,61}[0-9A-Za-z]$\" & !~ \" \""`
}
```
<a name="Owner"></a>
## type Owner {#Owner}
Owner represents the owner of a resource. For example, the name and email address of an engineering team.
```go
type Owner struct {
Name string
Email string
}
```
@@ -148,9 +269,40 @@ type Platform struct {
// Output represents the core Platform spec for the holos cli to iterate over
// and render each listed Component, injecting the Model.
Output core.Platform
// Domain represents the primary domain the Platform operates in. This field
// is intended as a sensible default for component authors to reference and
// platform operators to define.
Domain string `cue:"string | *\"holos.localhost\""`
}
```
<a name="Project"></a>
## type Project {#Project}
Project represents logical grouping of components owned by one or more teams. Useful for the platform team to manage resources for project teams to use.
```go
type Project struct {
// Name represents project name.
Name string
// Owner represents the team who own this project.
Owner Owner
// Namespaces represents the namespaces assigned to this project.
Namespaces map[core.NameLabel]Namespace
// Hostnames represents the host names to expose for this project.
Hostnames map[core.NameLabel]Hostname
}
```
<a name="Projects"></a>
## type Projects {#Projects}
Projects represents projects managed by the platform team for use by other teams using the platform.
```go
type Projects map[core.NameLabel]Project
```
<a name="StandardFleets"></a>
## type StandardFleets {#StandardFleets}
@@ -161,7 +313,7 @@ type StandardFleets struct {
// Workload represents a Fleet of zero or more workload Clusters.
Workload Fleet `json:"workload" cue:"{name: \"workload\"}"`
// Management represents a Fleet with one Cluster named management.
Management Fleet `json:"management" cue:"{name: \"management\", clusters: management: _}"`
Management Fleet `json:"management" cue:"{name: \"management\"}"`
}
```

View File

@@ -41,6 +41,7 @@ Note that Holos operates as a data pipeline, so the output of a [HelmChart](<#He
- [type Kustomize](<#Kustomize>)
- [type KustomizeBuild](<#KustomizeBuild>)
- [type Metadata](<#Metadata>)
- [type NameLabel](<#NameLabel>)
- [type Platform](<#Platform>)
- [type PlatformMetadata](<#PlatformMetadata>)
- [type PlatformSpec](<#PlatformSpec>)
@@ -328,6 +329,15 @@ type Metadata struct {
}
```
<a name="NameLabel"></a>
## type NameLabel {#NameLabel}
NameLabel is a unique identifier useful to convert a CUE struct to a list when the values have a Name field with a default value. This type is intended to indicate the common use case of converting a struct to a list where the Name field of the value aligns with the struct field name.
```go
type NameLabel string
```
<a name="Platform"></a>
## type Platform {#Platform}

407
doc/md/api/core/v1alpha4.md Normal file
View File

@@ -0,0 +1,407 @@
<!-- Code generated by gomarkdoc. DO NOT EDIT -->
# v1alpha4
```go
import "github.com/holos-run/holos/api/core/v1alpha4"
```
Package v1alpha4 contains the core API contract between the holos cli and CUE configuration code. Platform designers, operators, and software developers use this API to write configuration in CUE which \`holos\` loads. The overall shape of the API defines imperative actions \`holos\` should carry out to render the complete yaml that represents a Platform.
[Platform](<#Platform>) defines the complete configuration of a platform. With the holos reference platform this takes the shape of one management cluster and at least two workload clusters.
Each holos component path, e.g. \`components/namespaces\` produces exactly one [BuildPlan](<#BuildPlan>) which produces an [Artifact](<#Artifact>) collection. An [Artifact](<#Artifact>) is a fully rendered manifest produced from a [Transformer](<#Transformer>) sequence, which transforms a [Generator](<#Generator>) collection.
## Index
- [type Artifact](<#Artifact>)
- [type BuildPlan](<#BuildPlan>)
- [type BuildPlanSpec](<#BuildPlanSpec>)
- [type Chart](<#Chart>)
- [type Component](<#Component>)
- [type File](<#File>)
- [type FileContent](<#FileContent>)
- [type FileContentMap](<#FileContentMap>)
- [type FilePath](<#FilePath>)
- [type Generator](<#Generator>)
- [type Helm](<#Helm>)
- [type InternalLabel](<#InternalLabel>)
- [type Join](<#Join>)
- [type Kind](<#Kind>)
- [type Kustomization](<#Kustomization>)
- [type Kustomize](<#Kustomize>)
- [type Metadata](<#Metadata>)
- [type NameLabel](<#NameLabel>)
- [type Platform](<#Platform>)
- [type PlatformSpec](<#PlatformSpec>)
- [type Repository](<#Repository>)
- [type Resource](<#Resource>)
- [type Resources](<#Resources>)
- [type Transformer](<#Transformer>)
- [type Values](<#Values>)
<a name="Artifact"></a>
## type Artifact {#Artifact}
Artifact represents one fully rendered manifest produced by a [Transformer](<#Transformer>) sequence, which transforms a [Generator](<#Generator>) collection. A [BuildPlan](<#BuildPlan>) produces an [Artifact](<#Artifact>) collection.
Each Artifact produces one manifest file artifact. Generator Output values are used as Transformer Inputs. The Output field of the final [Transformer](<#Transformer>) should have the same value as the Artifact field.
When there is more than one [Generator](<#Generator>) there must be at least one [Transformer](<#Transformer>) to combine outputs into one Artifact. If there is a single Generator, it may directly produce the Artifact output.
An Artifact is processed concurrently with other artifacts in the same [BuildPlan](<#BuildPlan>). An Artifact should not use an output from another Artifact as an input. Each [Generator](<#Generator>) may also run concurrently. Each [Transformer](<#Transformer>) is executed sequentially starting after all generators have completed.
Output fields are write\-once. It is an error for multiple Generators or Transformers to produce the same Output value within the context of a [BuildPlan](<#BuildPlan>).
```go
type Artifact struct {
Artifact FilePath `json:"artifact,omitempty"`
Generators []Generator `json:"generators,omitempty"`
Transformers []Transformer `json:"transformers,omitempty"`
Skip bool `json:"skip,omitempty"`
}
```
<a name="BuildPlan"></a>
## type BuildPlan {#BuildPlan}
BuildPlan represents a build plan for holos to execute. Each [Platform](<#Platform>) component produces exactly one BuildPlan.
One or more [Artifact](<#Artifact>) files are produced by a BuildPlan, representing the fully rendered manifests for the Kubernetes API Server.
```go
type BuildPlan struct {
// Kind represents the type of the resource.
Kind string `json:"kind" cue:"\"BuildPlan\""`
// APIVersion represents the versioned schema of the resource.
APIVersion string `json:"apiVersion" cue:"string | *\"v1alpha4\""`
// Metadata represents data about the resource such as the Name.
Metadata Metadata `json:"metadata"`
// Spec specifies the desired state of the resource.
Spec BuildPlanSpec `json:"spec"`
}
```
<a name="BuildPlanSpec"></a>
## type BuildPlanSpec {#BuildPlanSpec}
BuildPlanSpec represents the specification of the [BuildPlan](<#BuildPlan>).
```go
type BuildPlanSpec struct {
// Component represents the component that produced the build plan.
// Represented as a path relative to the platform root.
Component string `json:"component"`
// Disabled causes the holos cli to disregard the build plan.
Disabled bool `json:"disabled,omitempty"`
// Artifacts represents the artifacts for holos to build.
Artifacts []Artifact `json:"artifacts"`
}
```
<a name="Chart"></a>
## type Chart {#Chart}
Chart represents a [Helm](<#Helm>) Chart.
```go
type Chart struct {
// Name represents the chart name.
Name string `json:"name"`
// Version represents the chart version.
Version string `json:"version"`
// Release represents the chart release when executing helm template.
Release string `json:"release"`
// Repository represents the repository to fetch the chart from.
Repository Repository `json:"repository,omitempty"`
}
```
<a name="Component"></a>
## type Component {#Component}
Component represents the complete context necessary to produce a [BuildPlan](<#BuildPlan>) from a [Platform](<#Platform>) component.
All of these fields are passed to the holos render component command using flags, which in turn are injected to CUE using tags. Field names should be used consistently through the platform rendering process for readability.
```go
type Component struct {
// Name represents the name of the component, injected as a tag to set the
// BuildPlan metadata.name field. Necessary for clear user feedback during
// platform rendering.
Name string `json:"name"`
// Component represents the path of the component relative to the platform root.
Component string `json:"component"`
// Cluster is the cluster name to provide when rendering the component.
Cluster string `json:"cluster"`
// Environment for example, dev, test, stage, prod
Environment string `json:"environment,omitempty"`
// Model represents the platform model holos gets from from the
// PlatformService.GetPlatform rpc method and provides to CUE using a tag.
Model map[string]any `json:"model"`
// Tags represents cue tags to inject when rendering the component. The json
// struct tag names of other fields in this struct are reserved tag names not
// to be used in the tags collection.
Tags []string `json:"tags,omitempty"`
}
```
<a name="File"></a>
## type File {#File}
File represents a simple single file copy [Generator](<#Generator>). Useful with a [Kustomize](<#Kustomize>) [Transformer](<#Transformer>) to process plain manifest files stored in the component directory. Multiple File generators may be used to transform multiple resources.
```go
type File struct {
// Source represents a file sub-path relative to the component path.
Source FilePath `json:"source"`
}
```
<a name="FileContent"></a>
## type FileContent {#FileContent}
FileContent represents file contents.
```go
type FileContent string
```
<a name="FileContentMap"></a>
## type FileContentMap {#FileContentMap}
FileContentMap represents a mapping of file paths to file contents.
```go
type FileContentMap map[FilePath]FileContent
```
<a name="FilePath"></a>
## type FilePath {#FilePath}
FilePath represents a file path.
```go
type FilePath string
```
<a name="Generator"></a>
## type Generator {#Generator}
Generator generates an intermediate manifest for a [Artifact](<#Artifact>).
Each Generator in a [Artifact](<#Artifact>) must have a distinct Output value for a [Transformer](<#Transformer>) to reference.
Refer to [Resources](<#Resources>), [Helm](<#Helm>), and [File](<#File>).
```go
type Generator struct {
// Kind represents the kind of generator. Must be Resources, Helm, or File.
Kind string `json:"kind" cue:"\"Resources\" | \"Helm\" | \"File\""`
// Output represents a file for a Transformer or Artifact to consume.
Output FilePath `json:"output"`
// Resources generator. Ignored unless kind is Resources. Resources are
// stored as a two level struct. The top level key is the Kind of resource,
// e.g. Namespace or Deployment. The second level key is an arbitrary
// InternalLabel. The third level is a map[string]any representing the
// Resource.
Resources Resources `json:"resources,omitempty"`
// Helm generator. Ignored unless kind is Helm.
Helm Helm `json:"helm,omitempty"`
// File generator. Ignored unless kind is File.
File File `json:"file,omitempty"`
}
```
<a name="Helm"></a>
## type Helm {#Helm}
Helm represents a [Chart](<#Chart>) manifest [Generator](<#Generator>).
```go
type Helm struct {
// Chart represents a helm chart to manage.
Chart Chart `json:"chart"`
// Values represents values for holos to marshal into values.yaml when
// rendering the chart.
Values Values `json:"values"`
// EnableHooks enables helm hooks when executing the `helm template` command.
EnableHooks bool `json:"enableHooks,omitempty"`
// Namespace represents the helm namespace flag
Namespace string `json:"namespace,omitempty"`
}
```
<a name="InternalLabel"></a>
## type InternalLabel {#InternalLabel}
InternalLabel is an arbitrary unique identifier internal to holos itself. The holos cli is expected to never write a InternalLabel value to rendered output files, therefore use a InternalLabel when the identifier must be unique and internal. Defined as a type for clarity and type checking.
```go
type InternalLabel string
```
<a name="Join"></a>
## type Join {#Join}
Join represents a [Join](<#Join>)\(https://pkg.go.dev/strings#Join\) [Transformer](<#Transformer>). Useful for the common case of combining the output of [Helm](<#Helm>) and [Resources](<#Resources>) [Generator](<#Generator>) into one [Artifact](<#Artifact>) when [Kustomize](<#Kustomize>) is otherwise unnecessary.
```go
type Join struct {
Separator string `json:"separator" cue:"string | *\"---\\n\""`
}
```
<a name="Kind"></a>
## type Kind {#Kind}
Kind is a discriminator. Defined as a type for clarity and type checking.
```go
type Kind string
```
<a name="Kustomization"></a>
## type Kustomization {#Kustomization}
Kustomization represents a kustomization.yaml file for use with the [Kustomize](<#Kustomize>) [Transformer](<#Transformer>). Untyped to avoid tightly coupling holos to kubectl versions which was a problem for the Flux maintainers. Type checking is expected to happen in CUE against the kubectl version the user prefers.
```go
type Kustomization map[string]any
```
<a name="Kustomize"></a>
## type Kustomize {#Kustomize}
Kustomize represents a kustomization [Transformer](<#Transformer>).
```go
type Kustomize struct {
// Kustomization represents the decoded kustomization.yaml file
Kustomization Kustomization `json:"kustomization"`
// Files holds file contents for kustomize, e.g. patch files.
Files FileContentMap `json:"files,omitempty"`
}
```
<a name="Metadata"></a>
## type Metadata {#Metadata}
Metadata represents data about the resource such as the Name.
```go
type Metadata struct {
// Name represents the resource name.
Name string `json:"name"`
}
```
<a name="NameLabel"></a>
## type NameLabel {#NameLabel}
NameLabel is a unique identifier useful to convert a CUE struct to a list when the values have a Name field with a default value. NameLabel indicates the common use case of converting a struct to a list where the Name field of the value aligns with the outer struct field name.
For example:
```
Outer: [NAME=_]: Name: NAME
```
```go
type NameLabel string
```
<a name="Platform"></a>
## type Platform {#Platform}
Platform represents a platform to manage. A Platform resource informs holos which components to build. The platform resource also acts as a container for the platform model form values provided by the PlatformService. The primary use case is to collect the cluster names, cluster types, platform model, and holos components to build into one resource.
```go
type Platform struct {
// Kind is a string value representing the resource.
Kind string `json:"kind" cue:"\"Platform\""`
// APIVersion represents the versioned schema of this resource.
APIVersion string `json:"apiVersion" cue:"string | *\"v1alpha4\""`
// Metadata represents data about the resource such as the Name.
Metadata Metadata `json:"metadata"`
// Spec represents the specification.
Spec PlatformSpec `json:"spec"`
}
```
<a name="PlatformSpec"></a>
## type PlatformSpec {#PlatformSpec}
PlatformSpec represents the specification of a [Platform](<#Platform>). Think of a platform spec as a [Component](<#Component>) collection for multiple kubernetes clusters combined with the user\-specified Platform Model.
```go
type PlatformSpec struct {
// Components represents a list of holos components to manage.
Components []Component `json:"components"`
}
```
<a name="Repository"></a>
## type Repository {#Repository}
Repository represents a [Helm](<#Helm>) [Chart](<#Chart>) repository.
```go
type Repository struct {
Name string `json:"name"`
URL string `json:"url"`
}
```
<a name="Resource"></a>
## type Resource {#Resource}
Resource represents one kubernetes api object.
```go
type Resource map[string]any
```
<a name="Resources"></a>
## type Resources {#Resources}
Resources represents a kubernetes resources [Generator](<#Generator>) from CUE.
```go
type Resources map[Kind]map[InternalLabel]Resource
```
<a name="Transformer"></a>
## type Transformer {#Transformer}
Transformer transforms [Generator](<#Generator>) manifests within a [Artifact](<#Artifact>).
```go
type Transformer struct {
// Kind represents the kind of transformer. Must be Kustomize, or Join.
Kind string `json:"kind" cue:"\"Kustomize\" | \"Join\""`
// Inputs represents the files to transform. The Output of prior Generators
// and Transformers.
Inputs []FilePath `json:"inputs"`
// Output represents a file for a subsequent Transformer or Artifact to
// consume.
Output FilePath `json:"output"`
// Kustomize transformer. Ignored unless kind is Kustomize.
Kustomize Kustomize `json:"kustomize,omitempty"`
// Join transformer. Ignored unless kind is Join.
Join Join `json:"join,omitempty"`
}
```
<a name="Values"></a>
## type Values {#Values}
Values represents [Helm](<#Helm>) Chart values generated from CUE.
```go
type Values map[string]any
```
Generated by [gomarkdoc](<https://github.com/princjef/gomarkdoc>)

View File

@@ -0,0 +1,724 @@
---
description: Try Holos with this quick start guide.
slug: /archive/2024-09-15-quickstart
sidebar_position: 100
---
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
import Admonition from '@theme/Admonition';
# Quickstart
In this guide, you'll experience how Holos makes the process of operating a
Platform safer, easier, and more consistent. We'll use Holos to manage a
vendor-provided Helm chart as a Component. Next, we'll mix in our own custom
resources to manage the Component with GitOps. Finally, you'll see how Holos
makes it safer and easier to maintain software over time by surfacing the exact
changes that will be applied when upgrading the vendor's chart to a new version,
before they are actually made.
The [Concepts](/docs/concepts) page defines capitalized terms such as Platform
and Component.
## What you'll need {#requirements}
You'll need the following tools installed to complete this guide.
1. [holos](/docs/install) - to build the Platform.
2. [helm](https://helm.sh/docs/intro/install/) - to render Holos Components that
wrap upstream Helm charts.
Optionally, if you'd like to apply the rendered manifests to a real Cluster,
first complete the [Local Cluster Guide](/docs/guides/local-cluster).
## Install Holos
Install Holos with the following command or other methods listed on the
[Installation](/docs/install/) page.
```bash
go install github.com/holos-run/holos/cmd/holos@latest
```
## Create a Git Repository
Start by initializing an empty Git repository. Holos operates on local files
stored in a Git repository.
<Tabs groupId="init">
<TabItem value="command" label="Command">
```bash
mkdir holos-quickstart
cd holos-quickstart
git init
```
</TabItem>
<TabItem value="output" label="Output">
```txt
Initialized empty Git repository in /holos-quickstart/.git/
```
</TabItem>
</Tabs>
This guide assumes you will run commands from the root directory of the Git
repository unless stated otherwise.
## Generate the Platform {#Generate-Platform}
Generate the Platform code in the repository root. A Platform refers to the
entire set of software holistically integrated to provide a software development
platform for your organization. In this guide, the Platform will include a
single Component to demonstrate how the concepts fit together.
```bash
holos generate platform quickstart
```
Commit the generated platform config to the repository.
<Tabs groupId="commit-platform">
<TabItem value="command" label="Command">
```bash
git add .
git commit -m "holos generate platform quickstart - $(holos --version)"
```
</TabItem>
<TabItem value="output" label="Output">
```txt
[main (root-commit) 0b17b7f] holos generate platform quickstart
213 files changed, 72349 insertions(+)
...
```
</TabItem>
</Tabs>
## Generate a Component {#generate-component}
The platform you generated is currently empty. Run the following command to
generate the CUE code that defines a Helm Component.
<Tabs groupId="gen-podinfo">
<TabItem value="command" label="Command">
```bash
holos generate component podinfo --component-version 6.6.1
```
</TabItem>
<TabItem value="output" label="Output">
```txt
generated component
```
</TabItem>
</Tabs>
The --component-version 6.6.1 flag intentionally installs an older release.
You'll see how Holos assists with software upgrades later in this guide.
The generate component command creates two files: a leaf file,
`components/podinfo/podinfo.gen.cue`, and a root file, `podinfo.gen.cue`. Holos
leverages the fact that [order is
irrelevant](https://cuelang.org/docs/tour/basics/order-irrelevance/) in CUE to
register the component with the Platform by adding a file to the root of the Git
repository. The second file defines the component in the leaf component
directory.
<Tabs groupId="podinfo-files">
<TabItem value="components/podinfo/podinfo.gen.cue" label="Leaf">
`components/podinfo/podinfo.gen.cue`
```cue showLineNumbers
package holos
// Produce a helm chart build plan.
(#Helm & Chart).Output
let Chart = {
Name: "podinfo"
Version: "6.6.1"
Namespace: "default"
Repo: name: "podinfo"
Repo: url: "https://stefanprodan.github.io/podinfo"
Values: {}
}
```
</TabItem>
<TabItem value="podinfo.gen.cue" label="Root">
`podinfo.gen.cue`
```cue showLineNumbers
package holos
// Manage podinfo on workload clusters only
for Cluster in #Fleets.workload.clusters {
#Platform: Components: "\(Cluster.name)/podinfo": {
path: "components/podinfo"
cluster: Cluster.name
}
}
```
</TabItem>
</Tabs>
In this example, we provide the minimal information needed to manage the Helm
chart: the name, version, Kubernetes namespace for deployment, and the chart
repository location.
This chart deploys cleanly without any values provided, but we include an empty
Values struct to show how Holos improves consistency and safety in Helm by
leveraging the strong type-checking in CUE. You can safely pass shared values,
such as the organizations domain name, to all Components across all clusters in
the Platform by defining them at the root of the configuration.
Commit the generated component config to the repository.
<Tabs groupId="commit-component">
<TabItem value="command" label="Command">
```bash
git add .
git commit -m "holos generate component podinfo - $(holos --version)"
```
</TabItem>
<TabItem value="output" label="Output">
```txt
[main cc0e90c] holos generate component podinfo
2 files changed, 24 insertions(+)
create mode 100644 components/podinfo/podinfo.gen.cue
create mode 100644 podinfo.gen.cue
```
</TabItem>
</Tabs>
## Render the Component
You can render individual components without adding them to a Platform, which is
helpful when developing a new component.
<Tabs groupId="render-podinfo">
<TabItem value="command" label="Command">
```bash
holos render component ./components/podinfo --cluster-name=default
```
</TabItem>
<TabItem value="output" label="Output">
```txt
cached
rendered podinfo
```
</TabItem>
</Tabs>
First, the command caches the Helm chart locally to speed up subsequent
renderings. Then, the command runs Helm to produce the output and writes it into
the deploy directory.
<Tabs groupId="tree-podinfo">
<TabItem value="command" label="Command">
```bash
tree deploy
```
</TabItem>
<TabItem value="output" label="Output">
```txt
deploy
└── clusters
└── default
└── components
└── podinfo
└── podinfo.gen.yaml
5 directories, 1 file
```
</TabItem>
</Tabs>
The component deploys to one cluster named `default`. In practice, the same
component is often deployed to multiple clusters, such as `east` and `west` to
provide redundancy and increase availability.
:::tip
This example is equivalent to running `helm template` on the chart and saving
the output to a file. Holos simplifies this task, making it safer and more
consistent when managing many charts.
:::
## Mix in an ArgoCD Application
We've seen how Holos works with Helm, but we haven't yet explored how Holos
makes it easier to consistently and safely manage all of the software in a
Platform.
Holos allows you to easily mix in resources that differentiate your Platform.
We'll use this feature to mix in an ArgoCD [Application][application] to manage
the podinfo Component with GitOps. We'll define this configuration in a way that
can be automatically and consistently reused across all future Components added
to the Platform.
Create a new file named `argocd.cue` in the root of your Git repository with the
following contents:
<Tabs groupId="argocd-config">
<TabItem value="command" label="argocd.cue">
```cue showLineNumbers
package holos
#ArgoConfig: {
Enabled: true
RepoURL: "https://github.com/holos-run/holos-quickstart-guide"
}
```
</TabItem>
</Tabs>
:::tip
If you plan to apply the rendered output to a real cluster, change the
`example.com` RepoURL to the URL of the Git repository you created in this
guide. You don't need to change the example if you're just exploring Holos by
inspecting the rendered output without applying it to a live cluster.
:::
With this file in place, render the component again.
<Tabs groupId="render-podinfo-argocd">
<TabItem value="command" label="Command">
```bash
holos render component ./components/podinfo --cluster-name=default
```
</TabItem>
<TabItem value="output" label="Output">
```txt
wrote deploy file
rendered gitops/podinfo
rendered podinfo
```
</TabItem>
</Tabs>
Holos uses the locally cached chart to improve performance and reliability. It
then renders the Helm template output along with an ArgoCD Application resource
for GitOps.
:::tip
By defining the ArgoCD configuration at the root, we again take advantage of the
fact that [order is
irrelevant](https://cuelang.org/docs/tour/basics/order-irrelevance/) in CUE.
:::
Defining the configuration at the root ensures all future leaf Components take
the ArgoCD configuration and render an Application manifest for GitOps
management.
<Tabs groupId="tree-podinfo-argocd">
<TabItem value="command" label="Command">
```bash
tree deploy
```
</TabItem>
<TabItem value="output" label="Output">
```txt
deploy
└── clusters
└── default
├── components
│   └── podinfo
│   └── podinfo.gen.yaml
└── gitops
└── podinfo.application.gen.yaml
6 directories, 2 files
```
</TabItem>
</Tabs>
Notice the new `podinfo.application.gen.yaml` file created by enabling ArgoCD in
the Helm component. The Application resource in the file looks like this:
<Tabs groupId="podinfo-application">
<TabItem value="file" label="podinfo.application.gen.yaml">
```yaml showLineNumbers
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
name: podinfo
namespace: argocd
spec:
destination:
server: https://kubernetes.default.svc
project: default
source:
path: ./deploy/clusters/default/components/podinfo
repoURL: https://example.com/holos-quickstart.git
targetRevision: main
```
</TabItem>
</Tabs>
:::tip
Holos generates a similar Application resource for every additional Component
added to your Platform.
:::
Finally, add and commit the results to your Platform's Git repository.
<Tabs groupId="commit-argo">
<TabItem value="command" label="Command">
```bash
git add .
git commit -m "holos render component ./components/podinfo --cluster-name=default"
```
</TabItem>
<TabItem value="output" label="Output">
```txt
[main f95cef1] holos render component ./components/podinfo --cluster-name=default
3 files changed, 134 insertions(+)
create mode 100644 argocd.cue
create mode 100644 deploy/clusters/default/components/podinfo/podinfo.gen.yaml
create mode 100644 deploy/clusters/default/gitops/podinfo.application.gen.yaml
```
</TabItem>
</Tabs>
In this section, we learned how Holos simplifies mixing resources into
Components, like an ArgoCD Application. Holos ensures consistency by managing an
Application resource for every Component added to the Platform through the
configuration you define in `argocd.cue` at the root of the repository.
## Define Workload Clusters {#workload-clusters}
We've generated a Component to manage podinfo and integrated it with our
Platform, but rendering the Platform doesn't render podinfo. Podinfo isn't
rendered because we haven't assigned any Clusters to the workload Fleet.
Define two new clusters, `east` and `west`, and assign them to the workload
Fleet. Create a new file named `clusters.cue` in the root of your Git repository
with the following contents:
<Tabs groupId="clusters">
<TabItem value="clusters.cue" label="clusters.cue">
```cue showLineNumbers
package holos
// Define two workload clusters for disaster recovery.
#Fleets: workload: clusters: {
// In CUE _ indicates values are defined elsewhere.
east: _
west: _
}
```
</TabItem>
</Tabs>
This example shows how Holos simplifies configuring multiple clusters with
similar configuration by grouping them into a Fleet.
:::tip
Fleets help segment a group of Clusters into one leader and multiple followers
by designating one cluster as the primary. Holos makes it safer, easier, and
more consistent to reconfigure which cluster is the primary. The primary can be
set to automatically restore persistent data from backups, while non-primary
clusters can be configured to automatically replicate from the primary.
Automatic database backup, restore, and streaming replication is an advanced
topic enabled by Cloud Native PG and CUE. Check back for a guide on this and
other Day 2 operations topics.
:::
## Render the Platform {#render-platform}
Render the Platform to render the podinfo Component for each of the workload
clusters.
<Tabs groupId="render-platform">
<TabItem value="command" label="Command">
```bash
holos render platform ./platform
```
</TabItem>
<TabItem value="output" label="Output">
```txt
rendered components/podinfo for cluster west in 99.480792ms
rendered components/podinfo for cluster east in 99.882667ms
```
</TabItem>
</Tabs>
The render platform command iterates over every Cluster in the Fleet and renders
each Component assigned to the Fleet. Notice the two additional subdirectories
created under the deploy directory, one for each cluster: `east` and `west`.
<Tabs groupId="tree-platform">
<TabItem value="command" label="Command">
```bash
tree deploy
```
</TabItem>
<TabItem value="output" label="Output">
```txt
deploy
└── clusters
├── default
│   ├── components
│   │   └── podinfo
│   │   └── podinfo.gen.yaml
│   └── gitops
│   └── podinfo.application.gen.yaml
# highlight-next-line
├── east
│   ├── components
│   │   └── podinfo
│   │   └── podinfo.gen.yaml
│   └── gitops
│   └── podinfo.application.gen.yaml
# highlight-next-line
└── west
├── components
│   └── podinfo
│   └── podinfo.gen.yaml
└── gitops
└── podinfo.application.gen.yaml
14 directories, 6 files
```
</TabItem>
</Tabs>
Holos ensures consistency and safety by defining the ArgoCD Application once,
with strong type checking, at the configuration root.
New Application resources are automatically generated for the `east` and `west`
workload Clusters.
<Tabs groupId="applications">
<TabItem value="east" label="east">
`deploy/clusters/east/gitops/podinfo.application.gen.yaml`
```yaml showLineNumbers
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
name: podinfo
namespace: argocd
spec:
destination:
server: https://kubernetes.default.svc
project: default
source:
# highlight-next-line
path: ./deploy/clusters/east/components/podinfo
repoURL: https://example.com/holos-quickstart.git
targetRevision: main
```
</TabItem>
<TabItem value="west" label="west">
`deploy/clusters/west/gitops/podinfo.application.gen.yaml`
```yaml showLineNumbers
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
name: podinfo
namespace: argocd
spec:
destination:
server: https://kubernetes.default.svc
project: default
source:
# highlight-next-line
path: ./deploy/clusters/west/components/podinfo
repoURL: https://example.com/holos-quickstart.git
targetRevision: main
```
</TabItem>
<TabItem value="default" label="default">
`deploy/clusters/default/gitops/podinfo.application.gen.yaml`
```yaml showLineNumbers
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
name: podinfo
namespace: argocd
spec:
destination:
server: https://kubernetes.default.svc
project: default
source:
# highlight-next-line
path: ./deploy/clusters/default/components/podinfo
repoURL: https://example.com/holos-quickstart.git
targetRevision: main
```
</TabItem>
</Tabs>
Add and commit the rendered Platform and workload Clusters.
<Tabs groupId="commit-render-platform">
<TabItem value="command" label="Command">
```bash
git add .
git commit -m "holos render platform ./platform - $(holos --version)"
```
</TabItem>
<TabItem value="output" label="Output">
```txt
[main 5aebcf5] holos render platform ./platform - 0.93.2
5 files changed, 263 insertions(+)
create mode 100644 clusters.cue
create mode 100644 deploy/clusters/east/components/podinfo/podinfo.gen.yaml
create mode 100644 deploy/clusters/east/gitops/podinfo.application.gen.yaml
create mode 100644 deploy/clusters/west/components/podinfo/podinfo.gen.yaml
create mode 100644 deploy/clusters/west/gitops/podinfo.application.gen.yaml
```
</TabItem>
</Tabs>
## Upgrade a Helm Chart
Holos is designed to ease the burden of Day 2 operations. With Holos, upgrading
software, integrating new software, and making safe platform-wide configuration
changes become easier.
Let's upgrade the podinfo Component to see how this works in practice. First,
update the Component version field to the latest upstream Helm chart version.
<Tabs groupId="gen-podinfo">
<TabItem value="command" label="Command">
```bash
holos generate component podinfo --component-version 6.6.2
```
</TabItem>
<TabItem value="output" label="Output">
```txt
generated component
```
</TabItem>
</Tabs>
Remove the cached chart version.
<Tabs groupId="gen-podinfo">
<TabItem value="command" label="Command">
```bash
rm -rf components/podinfo/vendor
```
</TabItem>
</Tabs>
Now re-render the Platform.
<Tabs groupId="render-platform2">
<TabItem value="command" label="Command">
```bash
holos render platform ./platform
```
</TabItem>
<TabItem value="output" label="Output">
```txt
rendered components/podinfo for cluster east in 327.10475ms
rendered components/podinfo for cluster west in 327.796541ms
```
</TabItem>
</Tabs>
Notice we're still using the upstream chart without modifying it. The Holos
component wraps around the chart to mix in additional resources and integrate
the component with the broader Platform.
## Visualize the Changes
Holos makes it easier to see exactly what changes are made and which resources
will be applied to the API server. By design, Holos operates on local files,
leaving the task of applying them to ecosystem tools like `kubectl` and ArgoCD.
This allows platform operators to inspect changes during code review, or before
committing the change at all.
For example, using `git diff`, we see that the only functional change when
upgrading this Helm chart is the deployment of a new container image tag to each
cluster. Additionally, we can roll out this change gradually by applying it to
the east cluster first, then to the west cluster, limiting the potential blast
radius of a problematic change.
<Tabs groupId="git-diff">
<TabItem value="command" label="Command">
```bash
git diff deploy/clusters/east
```
</TabItem>
<TabItem value="output" label="Output">
```diff showLineNumbers
diff --git a/deploy/clusters/east/components/podinfo/podinfo.gen.yaml b/deploy/clusters/east/components/podinfo/podinfo.gen.yaml
index 7cc3332..8c1647d 100644
--- a/deploy/clusters/east/components/podinfo/podinfo.gen.yaml
+++ b/deploy/clusters/east/components/podinfo/podinfo.gen.yaml
@@ -5,9 +5,9 @@ kind: Service
metadata:
name: podinfo
labels:
- helm.sh/chart: podinfo-6.6.1
+ helm.sh/chart: podinfo-6.6.2
app.kubernetes.io/name: podinfo
- app.kubernetes.io/version: "6.6.1"
+ app.kubernetes.io/version: "6.6.2"
app.kubernetes.io/managed-by: Helm
spec:
type: ClusterIP
@@ -29,9 +29,9 @@ kind: Deployment
metadata:
name: podinfo
labels:
- helm.sh/chart: podinfo-6.6.1
+ helm.sh/chart: podinfo-6.6.2
app.kubernetes.io/name: podinfo
- app.kubernetes.io/version: "6.6.1"
+ app.kubernetes.io/version: "6.6.2"
app.kubernetes.io/managed-by: Helm
spec:
replicas: 1
@@ -53,7 +53,7 @@ spec:
terminationGracePeriodSeconds: 30
containers:
- name: podinfo
# highlight-next-line
- image: "ghcr.io/stefanprodan/podinfo:6.6.1"
# highlight-next-line
+ image: "ghcr.io/stefanprodan/podinfo:6.6.2"
imagePullPolicy: IfNotPresent
command:
- ./podinfo
```
</TabItem>
</Tabs>
:::tip
Holos is designed to surface the _fully rendered_ manifests intended for the
Kubernetes API server, making it easier to see and reason about platform-wide
configuration changes.
:::
## Recap {#recap}
In this quickstart guide, we learned how Holos makes it easier, safer, and more
consistent to manage a Platform composed of multiple Clusters and upstream Helm
charts.
We covered how to:
1. Generate a Git repository for the Platform config.
2. Wrap the unmodified upstream podinfo Helm chart into a Component.
3. Render an individual Component.
4. Mix-in your Platform's unique resources to all Components. For example, ArgoCD Application resources.
5. Define multiple similar, but not identical, workload clusters.
6. Render the manifests for the entire Platform with the `holos render platform` command.
7. Upgrade a Helm chart to the latest version as an important Day 2 task.
8. Visualize and surface the details of planned changes Platform wide.
## Dive Deeper
If you'd like to dive deeper, check out the [Schema API][schema] and [Core
API][core] reference docs. The main difference between the schema and core
packages is that the schema is used by users to write refined CUE, while the
core package is what the schema produces for `holos` to execute. Users rarely
need to interact with the Core API when on the happy path, but can use the core
package as an escape hatch when the happy path doesn't go where you want.
[application]: https://argo-cd.readthedocs.io/en/stable/user-guide/application-specification/
[schema]: /docs/api/author/v1alpha3/
[core]: /docs/api/core/v1alpha3/

View File

@@ -0,0 +1,106 @@
---
description: Self service platform resource management for project teams.
slug: /archive/guides/2024-09-17-manage-a-project
sidebar_position: 250
---
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
import Admonition from '@theme/Admonition';
# Manage a Project
In this guide we'll explore how Holos easily, safely, and consistently manages
platform resources for teams to develop the projects they're working on.
Intended Audience: Platform Engineers and Software Engineers.
Goal is to demonstrate how the platform team can consistently, easily, and
safely provide platform resources to software engineers.
Assumption is software engineers have a container they want to deploy onto the
platform and make accessible. We'll use httpbin as a stand-in for the dev
team's container.
Project is roughly equivalent to Dev Team for the purpose of this guide, but in
practice multiple teams work on a given project over the lifetime of the
project, so we structure the files into projects instead of teams.
## What you'll need {#requirements}
You'll need the following tools installed to complete this guide.
1. [holos](/docs/install) - to build the Platform.
2. [helm](https://helm.sh/docs/intro/install/) - to render Helm Components.
3. [kubectl](https://kubernetes.io/docs/tasks/tools/) - to render Kustomize Components.
If you'd like to apply the manifests we render in this guide complete the
following optional, but recommended, steps.
a. Complete the [Local Cluster] guide to set up a local cluster to work with.
b. You'll need a GitHub account to fork the repository associated with this
guide.
## Fork the Guide Repository
<Tabs groupId="fork">
<TabItem value="command" label="Command">
```bash
```
</TabItem>
<TabItem value="output" label="Output">
```txt showLineNumbers
```
</TabItem>
</Tabs>
This guide assumes you will run commands from the root directory of this
repository unless stated otherwise.
[Quickstart]: /docs/quickstart
[Local Cluster]: /docs/guides/local-cluster
## Render the Platform
So we can build the basic platform. Don't dwell on the platform bits.
## Apply the Manifests
Deploy ArgoCD, but not any of the Application resources.
## Browse to ArgoCD
Note there is nothing here yet.
## Switch to your Fork
Note all of the Applications change consistently.
## Apply the Applications
Note how ArgoCD takes over management, no longer need to k apply.
## Create a Project
Project is a conceptual, not technical, thing in Holos. Mainly about how components are laid out in the filesystem tree.
We use a schematic built into holos as an example, the platform team could use the same or provide a similar template and instructions for development teams to self-serve.
## Render the Platform
Notice:
1. Project is registered with the platform at the root.
2. HTTPRoute and Namespace resources are added close to the root in `projects`
3. Deployment and Service resources are added at the leaf in `projects/httpbin/backend`
## Update the image tag
Add a basic schematic to demonstrate this. May need to add two new flags for image url and image tag to the generate subcommand, but should just be two new fields on the struct.
## Dive Deeper
Set the stage for constraints. Ideas: Limit what resources can be added,
namespaces can be operated in, enforce labels, etc...
Simple, consistent, easy constraints.

File diff suppressed because it is too large Load Diff

View File

@@ -2,4 +2,28 @@ import DocCardList from '@theme/DocCardList';
# Guides
## Technical Overview
Please see the [Technical Overview] to learn about Holos. If you're ready to
drive in and try Holos, please work through the following guides.
## Bank of Holos
The guides are organized as a progression. We'll use Holos to manage a
fictional bank's platform, the Bank of Holos in each of the guides. In doing so
we'll take the time to explain the foundational concepts of Holos.
1. [Quickstart] covers the foundational concepts of Holos.
2. [Deploy a Service] explains how to deploy a containerized service using an
existing Helm chart. This guide then explains how to deploy a similar service
safely and consistently with CUE instead of Helm.
3. [Change a Service] covers the day two task of making configuration changes to
deployed services safely and consistently.
---
<DocCardList />
[Quickstart]: /docs/quickstart/
[Deploy a Service]: /docs/guides/deploy-a-service/
[Change a Service]: /docs/guides/change-a-service/
[Technical Overview]: /docs/technical-overview/

View File

@@ -0,0 +1,714 @@
---
description: Change a service on your platform.
slug: /guides/change-a-service
sidebar_position: 300
---
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
import Admonition from '@theme/Admonition';
# Change a Service
In this guide, we'll explore how Holos supports the frontend development team at [Bank of Holos] in reconfiguring an already deployed service. Along the way, we'll demonstrate how simple configuration changes are made safer with type checking, and how rendering the complete platform provides clear visibility into those changes.
This guide builds on the concepts covered in the [Quickstart] and [Deploy a Service] guides.
## What you'll need {#requirements}
Like our other guides, this guide is intended to be useful without needing to
run each command. If you'd like to apply the manifests to a real Cluster,
complete the [Local Cluster Guide](/docs/guides/local-cluster) before this
guide.
You'll need the following tools installed to run the commands in this guide.
1. [holos](/docs/install) - to build the Platform.
2. [helm](https://helm.sh/docs/intro/install/) - to render Holos Components that
wrap Helm charts.
3. [kubectl](https://kubernetes.io/docs/tasks/tools/) - to render Holos
Components that render with Kustomize.
## Fork the Git Repository
If you haven't already done so, [fork the Bank of
Holos](https://github.com/holos-run/bank-of-holos/fork) then clone the
repository to your local machine.
<Tabs groupId="git-clone">
<TabItem value="command" label="Command">
```bash
# Change YourName
git clone https://github.com/YourName/bank-of-holos
cd bank-of-holos
```
</TabItem>
<TabItem value="output" label="Output">
```txt
Cloning into 'bank-of-holos'...
remote: Enumerating objects: 1177, done.
remote: Counting objects: 100% (1177/1177), done.
remote: Compressing objects: 100% (558/558), done.
remote: Total 1177 (delta 394), reused 1084 (delta 303), pack-reused 0 (from 0)
Receiving objects: 100% (1177/1177), 2.89 MiB | 6.07 MiB/s, done.
Resolving deltas: 100% (394/394), done.
```
</TabItem>
</Tabs>
Run the rest of the commands in this guide from the root of the repository.
If you plan to apply the changes we make, you can delete and re-create your
local platform synced to the start of this guide.
```bash
./scripts/reset-cluster
./scripts/apply
```
## Rename the Bank
Let's imagine the bank recently re-branded from The Bank of Holos to The
Holistic Bank. The software development team responsible for the front end
website needs to update the branding accordingly.
Let's explore how Holos catches errors early, before they land in production,
then guides the team to the best place to make a change.
The bank front end web service is managed by the
`projects/bank-of-holos/frontend/components/bank-frontend/` component which
refers to the organization display name in `schema.gen.cue`.
<Tabs groupId="F5B546EB-566F-4B83-84C3-C55B40F55555">
<TabItem value="schema.gen.cue" label="schema.gen.cue">
```cue showLineNumbers
package holos
import api "github.com/holos-run/holos/api/author/v1alpha3"
// Define the default organization name
// highlight-next-line
#Organization: DisplayName: string | *"Bank of Holos"
#Organization: Name: string | *"bank-of-holos"
#Organization: api.#OrganizationStrict
#Platform: api.#Platform
#Fleets: api.#StandardFleets
_ComponentConfig: {
Resources: #Resources
ArgoConfig: #ArgoConfig
}
#Helm: api.#Helm & _ComponentConfig
#Kustomize: api.#Kustomize & _ComponentConfig
#Kubernetes: api.#Kubernetes & _ComponentConfig
#ArgoConfig: api.#ArgoConfig & {
ClusterName: _ClusterName
}
```
</TabItem>
<TabItem value="projects/bank-of-holos/frontend/components/bank-frontend/bank-frontend.cue" label="projects/bank-of-holos/frontend/components/bank-frontend/bank-frontend.cue">
```cue showLineNumbers
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let Objects = {
Name: "bank-frontend"
Namespace: #BankOfHolos.Frontend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests/frontend.yaml
Resources: {
Service: frontend: {
metadata: name: "frontend"
metadata: labels: {
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
spec: {
selector: {
app: "frontend"
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
_ports: http: {
name: "http"
port: 80
targetPort: 8080
protocol: "TCP"
}
ports: [for x in _ports {x}]
}
}
Deployment: frontend: {
metadata: name: "frontend"
metadata: labels: {
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
spec: {
selector: matchLabels: {
app: "frontend"
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
template: {
metadata: labels: {
app: "frontend"
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
spec: {
securityContext: {
seccompProfile: type: "RuntimeDefault"
fsGroup: 1000
runAsGroup: 1000
runAsNonRoot: true
runAsUser: 1000
}
serviceAccountName: "bank-of-holos"
terminationGracePeriodSeconds: 5
containers: [{
env: [{
name: "BANK_NAME"
// highlight-next-line
value: #Organization.DisplayName
}, {
name: "ENV_PLATFORM"
value: "local"
}, {
name: "VERSION"
value: "v0.6.5"
}, {
name: "PORT"
value: "8080"
}, {
name: "ENABLE_TRACING"
value: "false"
}, {
name: "SCHEME"
value: "https"
}, {
name: "LOG_LEVEL"
value: "info"
}, {
name: "DEFAULT_USERNAME"
valueFrom: configMapKeyRef: {
key: "DEMO_LOGIN_USERNAME"
name: "demo-data-config"
}
}, {
name: "DEFAULT_PASSWORD"
valueFrom: configMapKeyRef: {
key: "DEMO_LOGIN_PASSWORD"
name: "demo-data-config"
}
}, {
name: "REGISTERED_OAUTH_CLIENT_ID"
valueFrom: configMapKeyRef: {
key: "DEMO_OAUTH_CLIENT_ID"
name: "oauth-config"
optional: true
}
}, {
name: "ALLOWED_OAUTH_REDIRECT_URI"
valueFrom: configMapKeyRef: {
key: "DEMO_OAUTH_REDIRECT_URI"
name: "oauth-config"
optional: true
}
}]
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "service-api-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/frontend:v0.6.5@sha256:d72050f70d12383e4434ad04d189b681dc625f696087ddf0b5df641645c9dafa"
livenessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 60
periodSeconds: 15
timeoutSeconds: 30
}
name: "front"
readinessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 10
periodSeconds: 5
timeoutSeconds: 10
}
resources: {
limits: {
cpu: "250m"
memory: "128Mi"
}
requests: {
cpu: "100m"
memory: "64Mi"
}
}
securityContext: {
allowPrivilegeEscalation: false
capabilities: drop: ["all"]
privileged: false
readOnlyRootFilesystem: true
}
volumeMounts: [{
mountPath: "/tmp"
name: "tmp"
}, {
mountPath: "/tmp/.ssh"
name: "publickey"
readOnly: true
}]
}]
volumes: [
{
emptyDir: {}
name: "tmp"
},
{
name: "publickey"
secret: {
items: [{key: "jwtRS256.key.pub", path: "publickey"}]
secretName: "jwt-key"
}
},
]
}
}
}
}
// Allow HTTPRoutes in the ingress gateway namespace to reference Services
// in this namespace.
ReferenceGrant: grant: #ReferenceGrant & {
metadata: namespace: Namespace
}
// Include shared resources
#BankOfHolos.Resources
}
}
```
</TabItem>
</Tabs>
Line 6 of the `schema.gen.cue` file defines the _default_ value for
`#Organization.DisplayName` by using `string | *"..."`. In CUE, the `*`
asterisk character denotes a [default value].
Line 78 of the `bank-frontend.cue` file refers to `#Organization.DisplayName` to
configure the front end web container.
Let's change the name of the bank by defining a new value for
`#Organization.DisplayName` at the root of the configuration. Create
`projects/organization.cue` with the following content.
<Tabs groupId="B386181F-EBE7-469D-8CB5-37631067669B">
<TabItem value="projects/organization.cue" label="projects/organization.cue">
```cue showLineNumbers
package holos
#Organization: DisplayName: "The Holistic-Bank"
```
</TabItem>
</Tabs>
Let's render the platform and see if this changes the name.
<Tabs groupId="A014333C-3271-4C22-87E6-2B7BF898EA3E">
<TabItem value="command" label="Command">
```bash
holos render platform ./platform
```
</TabItem>
<TabItem value="output" label="Output">
```txt
#Organization.DisplayName: 2 errors in empty disjunction:
#Organization.DisplayName: conflicting values "Bank of Holos" and "The Holistic-Bank":
/bank-of-holos/projects/organization.cue:3:29
/bank-of-holos/schema.gen.cue:6:39
// highlight-next-line
#Organization.DisplayName: invalid value "The Holistic-Bank" (out of bound =~"^[0-9A-Za-z][0-9A-Za-z ]{2,61}[0-9A-Za-z]$"):
/bank-of-holos/cue.mod/gen/github.com/holos-run/holos/api/author/v1alpha3/definitions_go_gen.cue:203:25
/bank-of-holos/cue.mod/gen/github.com/holos-run/holos/api/author/v1alpha3/definitions_go_gen.cue:188:15
/bank-of-holos/cue.mod/gen/github.com/holos-run/holos/api/author/v1alpha3/definitions_go_gen.cue:203:15
/bank-of-holos/projects/organization.cue:3:29
/bank-of-holos/schema.gen.cue:6:29
could not run: could not render component: exit status 1 at internal/render/platform.go:50
```
</TabItem>
</Tabs>
:::warning Whoops
The development team defined a value that isn't allowed by the
configuration.
:::
Someone else in the organization placed a [constraint] on the
configuration to ensure the display name contains only letters, numbers, and
spaces. This constraint is expressed as a [regular expression].
:::tip
CUE provides clear visibility where to start looking to resolve conflicts. Each
file and line number listed is a place the `#Organization.DisplayName` field is
defined.
:::
Let's try again, this time replacing the hyphen with a space.
<Tabs groupId="F93B34FA-C0C6-4793-A32F-DAD094403208">
<TabItem value="projects/organization.cue" label="projects/organization.cue">
```cue showLineNumbers
package holos
#Organization: DisplayName: "The Holistic Bank"
```
</TabItem>
</Tabs>
<Tabs groupId="5FD68778-476A-4F82-8817-71CEE205216E">
<TabItem value="command" label="Command">
```bash
holos render platform ./platform
```
</TabItem>
<TabItem value="output" label="Output">
```txt
rendered bank-ledger-db for cluster workload in 139.863625ms
rendered bank-accounts-db for cluster workload in 151.74875ms
rendered bank-balance-reader for cluster workload in 154.356083ms
rendered bank-ledger-writer for cluster workload in 161.209541ms
rendered bank-userservice for cluster workload in 163.373417ms
rendered bank-backend-config for cluster workload in 179.271208ms
rendered bank-secrets for cluster workload in 204.35625ms
rendered gateway for cluster workload in 118.707583ms
rendered httproutes for cluster workload in 140.981541ms
rendered bank-transaction-history for cluster workload in 156.066875ms
rendered bank-frontend for cluster workload in 300.102292ms
rendered bank-contacts for cluster workload in 159.89625ms
rendered cni for cluster workload in 150.754458ms
rendered istiod for cluster workload in 222.922625ms
rendered app-projects for cluster workload in 118.422792ms
rendered ztunnel for cluster workload in 142.840625ms
rendered cert-manager for cluster workload in 190.938834ms
rendered base for cluster workload in 340.679416ms
rendered local-ca for cluster workload in 107.120334ms
rendered external-secrets for cluster workload in 145.020834ms
rendered argocd for cluster workload in 299.690917ms
rendered namespaces for cluster workload in 115.862334ms
rendered gateway-api for cluster workload in 225.783833ms
rendered external-secrets-crds for cluster workload in 339.741166ms
rendered crds for cluster workload in 421.849041ms
rendered platform in 718.015959ms
```
</TabItem>
</Tabs>
:::tip Success
Great, the platform rendered. We know the display name is valid according to
the constraints.
:::
Let's see if the new display name value updated the configuration for the bank
frontend.
<Tabs groupId="6C068651-2061-4262-BE1E-7BB3E7EB66CB">
<TabItem value="command" label="Command">
```bash
git status
```
</TabItem>
<TabItem value="output" label="Output">
```txt
On branch main
Your branch and 'jeffmccune/main' have diverged,
and have 2 and 4 different commits each, respectively.
(use "git pull" to merge the remote branch into yours)
Changes not staged for commit:
(use "git add <file>..." to update what will be committed)
(use "git restore <file>..." to discard changes in working directory)
// highlight-next-line
modified: deploy/clusters/workload/components/app-projects/app-projects.gen.yaml
modified: deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml
Untracked files:
(use "git add <file>..." to include in what will be committed)
projects/organization.cue
no changes added to commit (use "git add" and/or "git commit -a")
```
</TabItem>
</Tabs>
<Tabs groupId="4A20831E-461B-4EDE-8F6E-E73C3AEC12DB">
<TabItem value="command" label="Command">
```bash
git diff
```
</TabItem>
<TabItem value="output" label="Output">
```diff
diff --git a/deploy/clusters/workload/components/app-projects/app-projects.gen.yaml b/deploy/clusters/workload/components/app-projects/app-projects.gen.yaml
index 7914756..250c660 100644
--- a/deploy/clusters/workload/components/app-projects/app-projects.gen.yaml
+++ b/deploy/clusters/workload/components/app-projects/app-projects.gen.yaml
@@ -9,7 +9,7 @@ spec:
clusterResourceWhitelist:
- group: '*'
kind: '*'
- description: Holos managed AppProject for Bank of Holos
+ description: Holos managed AppProject for The Holistic Bank
destinations:
- namespace: '*'
server: '*'
@@ -26,7 +26,7 @@ spec:
clusterResourceWhitelist:
- group: '*'
kind: '*'
- description: Holos managed AppProject for Bank of Holos
+ description: Holos managed AppProject for The Holistic Bank
destinations:
- namespace: '*'
server: '*'
@@ -43,7 +43,7 @@ spec:
clusterResourceWhitelist:
- group: '*'
kind: '*'
- description: Holos managed AppProject for Bank of Holos
+ description: Holos managed AppProject for The Holistic Bank
destinations:
- namespace: '*'
server: '*'
@@ -60,7 +60,7 @@ spec:
clusterResourceWhitelist:
- group: '*'
kind: '*'
- description: Holos managed AppProject for Bank of Holos
+ description: Holos managed AppProject for The Holistic Bank
destinations:
- namespace: '*'
server: '*'
diff --git a/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml b/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml
index dae6f93..d41516b 100644
--- a/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml
+++ b/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml
@@ -71,7 +71,7 @@ spec:
containers:
- env:
- name: BANK_NAME
- value: Bank of Holos
+ value: The Holistic Bank
- name: ENV_PLATFORM
value: local
- name: VERSION
```
</TabItem>
</Tabs>
:::danger
The new display name changed the frontend container, but it _also_ affected the
app-projects component owned by the platform team.
:::
Submitting a pull request would trigger a code review from the platform
engineering team who manages the app-projects component. Let's see how to
narrow the change down to limit the scope to the bank's user facing services.
All of these services are managed under `projects/bank-of-holos/` Move the
`organization.cue` file into this folder to limit the scope of configuration to
the the components contained within.
```bash
mv projects/organization.cue projects/bank-of-holos/
```
Render the platform and let's see what changed.
<Tabs groupId="0FFEC244-B59B-4136-9C82-837985DC2AB8">
<TabItem value="command" label="Command">
```bash
holos render platform ./platform
```
</TabItem>
<TabItem value="output" label="Output">
```txt
rendered bank-ledger-db for cluster workload in 163.814917ms
rendered bank-accounts-db for cluster workload in 163.960208ms
rendered bank-userservice for cluster workload in 164.1625ms
rendered bank-ledger-writer for cluster workload in 169.185291ms
rendered bank-balance-reader for cluster workload in 174.5455ms
rendered bank-backend-config for cluster workload in 178.092125ms
rendered bank-secrets for cluster workload in 202.305334ms
rendered gateway for cluster workload in 122.81725ms
rendered httproutes for cluster workload in 134.121084ms
rendered bank-contacts for cluster workload in 146.4185ms
rendered bank-frontend for cluster workload in 311.35425ms
rendered bank-transaction-history for cluster workload in 160.103ms
rendered cni for cluster workload in 145.762083ms
rendered istiod for cluster workload in 216.0065ms
rendered app-projects for cluster workload in 117.684333ms
rendered ztunnel for cluster workload in 144.555292ms
rendered cert-manager for cluster workload in 178.247917ms
rendered base for cluster workload in 336.679ms
rendered external-secrets for cluster workload in 142.21825ms
rendered local-ca for cluster workload in 101.249ms
rendered argocd for cluster workload in 280.54525ms
rendered namespaces for cluster workload in 106.822042ms
rendered gateway-api for cluster workload in 200.459791ms
rendered external-secrets-crds for cluster workload in 470.125833ms
rendered crds for cluster workload in 844.388666ms
rendered platform in 1.154937084s
```
</TabItem>
</Tabs>
<Tabs groupId="DE4FEEE5-FC53-48A6-BC6F-D0EA1DBFD00C">
<TabItem value="command" label="Command">
```bash
git diff
```
</TabItem>
<TabItem value="output" label="Output">
```diff
diff --git a/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml b/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml
index dae6f93..d41516b 100644
--- a/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml
+++ b/deploy/clusters/workload/components/bank-frontend/bank-frontend.gen.yaml
@@ -71,7 +71,7 @@ spec:
containers:
- env:
- name: BANK_NAME
- value: Bank of Holos
+ value: The Holistic Bank
- name: ENV_PLATFORM
value: local
- name: VERSION
```
</TabItem>
</Tabs>
:::tip Success
Great! This time, the only manifest affected is our `bank-frontend.gen.yaml`.
:::
The `BANK_NAME` environment variable will change as we expect, and only the dev
teams managing the bank services components are affected by the change.
Let's commit and push this change and see if it works.
<Tabs groupId="435D9C60-F841-4CF1-A947-506422E6BAC9">
<TabItem value="command" label="Command">
```bash
git add .
git commit -m 'frontend: rename bank to The Holistic Bank'
git push
```
</TabItem>
<TabItem value="output" label="Output">
```txt
[main fda74ec] frontend: rename bank to The Holistic Bank
2 files changed, 4 insertions(+), 1 deletion(-)
create mode 100644 projects/bank-of-holos/organization.cue
```
</TabItem>
</Tabs>
Now that we've pushed the change, let's apply the change to the platform.
## Apply the Change
Once we've pushed the change, navigate to the [bank-frontend GitOps
Application](https://argocd.holos.localhost/applications/argocd/bank-frontend?view=tree&resource=).
We can see the Deployment needs to sync to the desired state we just pushed.
![bank-frontend out of sync](./img/change-a-service-out-of-sync.png)
Clicking on the frontend Deployment, we see the diff with the change we expect.
![bank-frontend diff](./img/change-a-service-diff.png)
Sync the change, ArgoCD applies the desired configuration state to the cluster
and Kubernetes handles rolling out the updated Deployment resource.
![bank-frontend progressing](./img/change-a-service-progressing.png)
Soon, the deployment finishes and the component is in sync again.
![bank-frontend in sync](./img/change-a-service-in-sync.png)
Finally, let's see if the name actually changed on the website. Navigate to
https://bank.holos.localhost/.
![bank-frontend login page](./img/change-a-service-login-page.png)
:::tip Success
We successfully made our change and successfully applied the changed
configuration to the platform.
:::
Thanks for taking the time to work through this guide which covered:
- How multiple teams could be impacted by defining configuration at the
`projects/` path.
- How to scope our change to only affect components within the
`projects/bank-of-holos/` path, eliminating the impact on other teams.
- How CUE can [constrain] values in Holos, increasing safety.
- How to handle a [default value] in CUE.
- How CUE surfaces the file and line number of _every_ place to look for where a
value is defined, making it faster and easier to troubleshoot problems.
[Quickstart]: /docs/quickstart/
[Deploy a Service]: /docs/guides/deploy-a-service/
[Change a Service]: /docs/guides/change-a-service/
[Helm]: /docs/api/author/v1alpha3/#Helm
[Kubernetes]: /docs/api/author/v1alpha3/#Kubernetes
[Kustomize]: /docs/api/author/v1alpha3/#Kustomize
[ComponentFields]: /docs/api/author/v1alpha3/#ComponentFields
[platform-files]: /docs/quickstart/#how-platform-rendering-works
[AppProject]: https://argo-cd.readthedocs.io/en/stable/user-guide/projects/
[unification operator]: https://cuelang.org/docs/reference/spec/#unification
[code-owners]: https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/about-code-owners
[Kustomization API]: https://github.com/kubernetes-sigs/kustomize/blob/release-kustomize-v5.2/api/types/kustomization.go#L34
[cue import]: https://cuelang.org/docs/reference/command/cue-help-import/
[cue get go]: https://cuelang.org/docs/concept/how-cue-works-with-go/
[timoni-crds]: https://timoni.sh/cue/module/custom-resources/
[HTTPRoute]: https://gateway-api.sigs.k8s.io/api-types/httproute/?h=filter
[Ingress]: https://kubernetes.io/docs/concepts/services-networking/ingress/
[hidden field]: https://cuelang.org/docs/tour/references/hidden/
[comprehension]: https://cuelang.org/docs/reference/spec/#comprehensions
[code owners]: https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/about-code-owners
[ReferenceGrant]: https://gateway-api.sigs.k8s.io/api-types/referencegrant/
[Local Cluster Guide]: /docs/guides/local-cluster
[Bank of Holos]: https://github.com/holos-run/bank-of-holos
[default value]: https://cuelang.org/docs/tour/types/defaults/
[constrain]: https://cuelang.org/docs/tour/basics/constraints/
[constraint]: https://cuelang.org/docs/tour/basics/constraints/
[regular expression]: https://cuelang.org/docs/tour/expressions/regexp/

File diff suppressed because it is too large Load Diff

View File

@@ -1,95 +0,0 @@
---
description: Use Holos to expose a Service with the Gateway API.
slug: /guides/expose-a-service
sidebar_position: 300
---
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
import Admonition from '@theme/Admonition';
# Expose a Service
In this guide, you'll learn how to expose a service with Holos using the Gateway
API.
:::warning TODO
Complete this section once the steps are complete.
:::
The [Concepts](/docs/concepts) page defines capitalized terms such as Platform
and Component.
## What you'll need {#requirements}
:::warning TODO
Complete this section once the steps are complete.
:::
You'll need the following tools installed to complete this guide.
1. [holos](/docs/install) - to build the Platform.
2. [helm](https://helm.sh/docs/intro/install/) - to render Holos Components that
wrap upstream Helm charts.
Optionally, if you'd like to apply the rendered manifests to a real Cluster,
first complete the [localhost Guide](../local-cluster).
## Create a Git Repository
Start by initializing an empty Git repository. Holos operates on local files
stored in a Git repository.
<Tabs groupId="init">
<TabItem value="command" label="Command">
```bash
mkdir expose-a-service
cd expose-a-service
git init
```
</TabItem>
<TabItem value="output" label="Output">
```txt
Initialized empty Git repository in /expose-a-service/.git/
```
</TabItem>
</Tabs>
This guide assumes you will run commands from the root directory of the Git
repository unless stated otherwise.
## Generate the Platform {#Generate-Platform}
Start by generating a platform used as the basis for our guides.
```bash
holos generate platform guide
```
Commit the generated platform config to the repository.
<Tabs groupId="commit-platform">
<TabItem value="command" label="Command">
```bash
git add .
git commit -m "holos generate platform guide - $(holos --version)"
```
</TabItem>
<TabItem value="output" label="Output">
```txt
[main (root-commit) 0b17b7f] holos generate platform guide - 0.93.3
213 files changed, 72349 insertions(+)
...
```
</TabItem>
</Tabs>
## Manage httpbin {#manage-httpbin}
The platform you generated is currently empty. Run the following command to
generate a Holos Component for the
[httpbin](https://github.com/mccutchen/go-httpbin) service.
httpbin is a simple backend service useful for end-to-end testing. In this
guide, we use httpbin as a example of a service your organization develops and
deploy onto your Platform.

View File

@@ -0,0 +1,15 @@
---
description: Helm Component
slug: /guides/helm-component
sidebar_position: 400
---
# Helm Component
The [Deploy a Service](/docs/guides/deploy-a-service/) guide is the best guide
we have on wrapping a Helm chart in a Holos Component. The [Helm] section of
the Author API may also be useful.
[Helm]: /docs/api/author/v1alpha3/#Helm
[Kubernetes]: /docs/api/author/v1alpha3/#Kubernetes
[Kustomize]: /docs/api/author/v1alpha3/#Kustomize

Binary file not shown.

After

Width:  |  Height:  |  Size: 690 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 997 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.1 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 287 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.1 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.1 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1009 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 617 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 706 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 794 KiB

View File

@@ -0,0 +1,20 @@
---
description: Kubernetes Component
slug: /guides/kubernetes-component
sidebar_position: 500
---
# Kubernetes Component
:::warning
TODO
:::
This is a placeholder for a guide for managing Kubernetes resources directly
from a Holos Component with strong type checking.
In the meantime, please refer to the [Kubernetes] section of the Author API.
[Helm]: /docs/api/author/v1alpha3/#Helm
[Kubernetes]: /docs/api/author/v1alpha3/#Kubernetes
[Kustomize]: /docs/api/author/v1alpha3/#Kustomize

View File

@@ -0,0 +1,20 @@
---
description: Wrap a Kustomize Kustomization in a Holos Component.
slug: /guides/kustomize-component
sidebar_position: 600
---
# Kustomize Component
:::warning
TODO
:::
This is a placeholder for a guide on wrapping a Kustomize Kustomization base
with a Holos component.
In the meantime, please refer to the [Kustomize] section of the Author API.
[Helm]: /docs/api/author/v1alpha3/#Helm
[Kubernetes]: /docs/api/author/v1alpha3/#Kubernetes
[Kustomize]: /docs/api/author/v1alpha3/#Kustomize

View File

@@ -1,7 +1,7 @@
---
description: Build a local Cluster to use with these guides.
slug: /guides/local-cluster
sidebar_position: 200
sidebar_position: 999
---
import Tabs from '@theme/Tabs';
@@ -10,15 +10,97 @@ import Admonition from '@theme/Admonition';
# Local Cluster
In this guide you'll set up a Cluster on your local host to apply and explore
the configuration described in our other guides. After completing this guide
you'll have a standard Kubernetes API server with proper DNS and TLS
certificates. You'll be able to easily reset the cluster to a known good state
to iterate on your own Platform.
In this guide we'll set up a local k3d cluster to apply and explore the
configuration described in our other guides. After completing this guide you'll
have a standard Kubernetes API server with proper DNS and TLS certificates.
You'll be able to easily reset the cluster to a known good state to iterate on
your own Platform.
The [Concepts](/docs/concepts) page defines capitalized terms such as Platform
and Component.
## Reset the Cluster
If you've already followed this guide, reset the cluster by running the
following commands. Skip this section if you're creating a cluster for the
first time.
First, delete the cluster.
<Tabs groupId="k3d-cluster-delete">
<TabItem value="command" label="Command">
```bash
k3d cluster delete workload
```
</TabItem>
<TabItem value="output" label="Output">
```txt showLineNumbers
INFO[0000] Deleting cluster 'workload'
INFO[0000] Deleting cluster network 'k3d-workload'
INFO[0000] Deleting 1 attached volumes...
INFO[0000] Removing cluster details from default kubeconfig...
INFO[0000] Removing standalone kubeconfig file (if there is one)...
INFO[0000] Successfully deleted cluster workload!
```
</TabItem>
</Tabs>
Then create the cluster again.
<Tabs groupId="k3d-cluster-create">
<TabItem value="command" label="Command">
```bash
k3d cluster create workload \
--registry-use k3d-registry.holos.localhost:5100 \
--port "443:443@loadbalancer" \
--k3s-arg "--disable=traefik@server:0"
```
</TabItem>
<TabItem value="output" label="Output">
```txt showLineNumbers
INFO[0000] portmapping '443:443' targets the loadbalancer: defaulting to [servers:*:proxy agents:*:proxy]
INFO[0000] Prep: Network
INFO[0000] Created network 'k3d-workload'
INFO[0000] Created image volume k3d-workload-images
INFO[0000] Starting new tools node...
INFO[0000] Starting node 'k3d-workload-tools'
INFO[0001] Creating node 'k3d-workload-server-0'
INFO[0001] Creating LoadBalancer 'k3d-workload-serverlb'
INFO[0001] Using the k3d-tools node to gather environment information
INFO[0001] HostIP: using network gateway 172.17.0.1 address
INFO[0001] Starting cluster 'workload'
INFO[0001] Starting servers...
INFO[0001] Starting node 'k3d-workload-server-0'
INFO[0003] All agents already running.
INFO[0003] Starting helpers...
INFO[0003] Starting node 'k3d-workload-serverlb'
INFO[0009] Injecting records for hostAliases (incl. host.k3d.internal) and for 3 network members into CoreDNS configmap...
INFO[0012] Cluster 'workload' created successfully!
INFO[0012] You can now use it like this:
kubectl cluster-info
```
</TabItem>
</Tabs>
Finally, add your trusted certificate authority.
<Tabs groupId="apply-local-ca">
<TabItem value="command" label="Command">
```bash
kubectl apply --server-side=true -f "$(mkcert -CAROOT)/namespace.yaml"
kubectl apply --server-side=true -n cert-manager -f "$(mkcert -CAROOT)/local-ca.yaml"
```
</TabItem>
<TabItem value="output" label="Output">
```txt showLineNumbers
namespace/cert-manager serverside-applied
secret/local-ca serverside-applied
```
</TabItem>
</Tabs>
You're back to the same state as the first time you completed this guide.
## What you'll need {#requirements}
You'll need the following tools installed to complete this guide.
@@ -150,10 +232,11 @@ cp -p "${CAROOT}/rootCA.pem" ca.crt
cp -p "${CAROOT}/rootCA.pem" tls.crt
cp -p "${CAROOT}/rootCA-key.pem" tls.key
kubectl create secret generic --from-file=. --dry-run=client -o yaml local-ca > ../local-ca.yaml
echo 'type: kubernetes.io/tls' >> ../local-ca.yaml
cd ..
echo 'type: kubernetes.io/tls' >> local-ca.yaml
kubectl apply --server-side=true -f- <<EOF
cat <<EOF > namespace.yaml
apiVersion: v1
kind: Namespace
metadata:
@@ -164,8 +247,12 @@ spec:
finalizers:
- kubernetes
EOF
kubectl apply --server-side=true -f namespace.yaml
kubectl apply -n cert-manager --server-side=true -f local-ca.yaml
# Save the Secret to easily reset the cluster later.
install -m 0644 namespace.yaml "${CAROOT}/namespace.yaml"
install -m 0600 local-ca.yaml "${CAROOT}/local-ca.yaml"
```
:::warning
@@ -184,12 +271,6 @@ with:
k3d cluster delete workload
```
## Reset {#reset}
If you'd like to reset to a known good state, execute the [Clean Up](#clean-up)
section, then [Create the Cluster](#create-the-cluster) and the [Setup Root
CA](#setup-root-ca) tasks.
## Next Steps
Now that you have a real cluster, apply and explore the manifests Holos renders

File diff suppressed because it is too large Load Diff

View File

@@ -1,64 +0,0 @@
# Introduction
⚡️ Holos will help you build your **software development platform in no time.**
💸 Building a software development platform is **time consuming and expensive**. Spend more time building features for your customers and less time managing your development platform.
💥 Already have a platform? Add new features and services to your platform easily with Holos.
🧐 Holos is a platform builder. It builds a hollistic software development platform composed of best-of-breed cloud native open source projects. Holos is also a tool to make it easier to manage cloud infrastructure by providing a typed alternative to yaml templates.
## Features
Holos was built to solve two main problems:
1. Building a platform usually takes 3 engineers 6-9 months of effort. Holos provides a reference platform that enables you to deploy and customize your platform in a fraction of the time.
2. Configuration changes often cause outages. Existing tools like Helm make it difficult to understand the impact a configuration change will have. Holos provides a unique, unified configuration model powered by CUE that makes it safer and easier to roll out configuration changes.
A core principle of Holos is that organizations gain value from owning the the platform they build on. Avoid vendor lock-in, future price hikes, and expensive licensing changes by building on a solid foundation of open source, cloud native computing foundation backed projects.
The following features are built into the Holos reference platform.
:::tip
Don't see your preferred technology in the stack? Holos is designed to enable you to swap out components of the platform tech stack.
:::
- **Continuous Delivery**
- Holos builds a GitOps workflow for each application running in the platform.
- Developers push changes which are automatically deployed.
- Powered by [ArgoCD](https://argo-cd.readthedocs.io/en/stable/)
- **Identity and Access Management** (IAM)
- Holos builds a standard OIDC identity provider for you.
- Integrates with your exisitng IAM and SSO system, or works independently.
- Powerful customer identity and access management features.
- Role based access control.
- Powered by [ZITADEL](https://zitadel.com/)
- **Zero Trust**
- Authenticate and Authorize users at the platform layer instead of or in addition to the application layer.
- Integrated with observability to measure and alert about problems before customers complain.
- Powered by [Istio](https://istio.io/)
- **Observability**
- Holos collects performance and availability metrics automatically, without requiring application changes.
- Optional, deeper integration into the application layer.
- Distributed Tracing
- Logging
- Powered by Prometheus, Grafana, Loki, and OpenTelemetry.
- **Data Platform**
- Integrated management of PostgreSQL
- Automatic backups
- Automatic restore from backup
- Quickly fail over across multiple regions
- **Multi-Region**
- Holos is designed to operate in multiple regions and multiple clouds.
- Keep customer data in the region that makes the most sense for your business.
- Easily cut over from one region to another for redundancy and business continuity.
## Development Status
Holos is being actively developed by [Open Infrastructure Services](https://openinfrastructure.co). Release can be found [here](https://github.com/holos-run/holos/releases).
## Adoption
Organizations who have officially adopted Holos can be found [here](https://github.com/holos-run/holos/blob/main/ADOPTERS.md).

View File

@@ -5,6 +5,49 @@ slug: /
# Introduction
:::warning TODO
See [introduction](https://github.com/facebook/docusaurus/blob/main/website/docs/introduction.mdx?plain=1)
:::
Welcome to Holos. Holos is an open source tool to manage software development
platforms safely, easily, and consistently. We built Holos to help engineering
teams work more efficiently together by empowering them to build golden paths
and paved roads for other teams to leverage for quicker delivery.
## Documentation
- [Guides] are organized into example use-cases of how Holos helps engineering
teams at the fictional Bank of Holos deliver business value on the bank's
platform.
- The [API Reference] is a technical reference for when you're writing CUE code to define your platform.
## Backstory
At [Open Infrastructure Services], we've each helped dozens of companies build and operate their software development platforms. During the U.S. presidential election just before the pandemic, our second-largest client, Twitter, experienced a global outage that lasted nearly a full day. We were managing their production configuration system, allowing the core infrastructure team to focus on business-critical objectives. This gave us a front-row seat to the incident.
A close friend and engineer on the team made a trivial one-line change to the firewall configuration. Less than 30 minutes later, everything was down. That change, which passed code review, caused the host firewall to revert to its default state on hundreds of thousands of servers, blocking all connections globally—except for SSH, thankfully. Even a Presidential candidate complained loudly.
This incident forced us to reconsider key issues with Twitter's platform:
1. **Lack of Visibility** - Engineers couldn't foresee the impact of even a small change, making it difficult to assess risks.
2. **Large Blast Radius** - Small changes affected the entire global fleet in under 30 minutes. There was no way to limit the impact of a single change.
3. **Incomplete Tooling** - The right processes were in place, but the tooling didn't fully support them. The change was tested and reviewed, but critical information wasn't surfaced in time.
Over the next few years, we built features to address these issues. Meanwhile, I began exploring how these solutions could work in the Kubernetes and cloud-native space.
As Google Cloud partners, we worked with large customers to understand how they built their platforms on Kubernetes. During the pandemic, we built a platform using CNCF projects like ArgoCD, Prometheus Stack, Istio, Cert Manager, and External Secrets Operator, integrating them into a cohesive platform. We started with upstream recommendations—primarily Helm charts—and wrote scripts to integrate each piece into the platform. For example, we passed Helm outputs to Kustomize to add labels or fix bugs, and wrote umbrella charts to add Ingress, HTTPRoute, and ExternalSecret resources.
These scripts served as necessary glue to hold everything together but became difficult to manage across multiple environments, regions, and cloud providers. YAML templates and nested loops created friction, making them hard to troubleshoot. The scripts themselves made it difficult to see what was happening and to fix issues affecting the entire platform.
Still, the scripts had a key advantage: they produced fully rendered manifests in plain text, committed to version control, and applied via ArgoCD. This clarity made troubleshooting easier and reduced errors in production.
Despite the makeshift nature of the scripts, I kept thinking about the "[Why are we templating YAML]?" post on Hacker News. I wanted to replace our scripts and charts with something more robust and easier to maintain—something that addressed Twitter's issues head-on.
I rewrote our scripts and charts using CUE and Go, replacing the glue layer. The result is **Holos**—a tool designed to complement Helm, Kustomize, and Jsonnet, making it easier and safer to define golden paths and paved roads without bespoke scripts or templates.
Thanks for reading. Take Holos for a spin on your local machine with our [Quickstart] guide.
[Guides]: /docs/guides/
[API Reference]: /docs/api/
[Quickstart]: /docs/quickstart/
[CUE]: https://cuelang.org/
[Author API]: /docs/api/author/
[Core API]: /docs/api/core/
[Open Infrastructure Services]: https://openinfrastructure.co/
[Why are we templating YAML]: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&query=https%3A%2F%2Fleebriggs.co.uk%2Fblog%2F2019%2F02%2F07%2Fwhy-are-we-templating-yaml&sort=byDate&type=story

View File

@@ -2,4 +2,12 @@ import DocCardList from '@theme/DocCardList';
# Get Started
## Start with the [Quickstart] guide
---
These documents provide additional context to supplement the [Quickstart] guide.
<DocCardList />
[Quickstart]: /docs/quickstart/

View File

@@ -6,65 +6,10 @@ sidebar_position: 300
# Comparison
:::tip
Holos is designed to complement and improve, not replace, existing tools in the
cloud native ecosystem.
:::
## Helm
### Chart Users
Describe how things are different when using an upstream helm chart.
### Chart Authors
Describe how things are different when writing a new helm chart.
## Kustomize
TODO
## ArgoCD
TODO
## Flux
TODO
## Timoni
| Aspect | Timoni | Holos | Comment |
| ---------- | -------------------- | -------------------- | ---------------------------------------------------------------------------------------- |
| Language | CUE | CUE | Like Holos, Timoni is also built on CUE. |
| Artifact | OCI Image | Plain YAML Files | The Holos Authors find plain files easier to work with and reason about than OCI images. |
| Outputs to | OCI Image Repository | Local Git repository | Holos is designed for use with existing GitOps tools. |
| Concept | Module | Component | A Timoni Module is analogous to a Holos Component. |
| Concept | Bundle | Platform | A Timoni Bundle is somewhat similar, but smaller in scope to a Holos Platform. |
:::important
The Holos Authors are deeply grateful to Stefan and Timoni for the capability of
importing Kubernetes custom resource definitions into CUE. Without this
functionality, much of the Kubernetes ecosystem would be more difficult to
manage in CUE and therefore in Holos.
:::
## KubeVela
1. Also built on CUE.
2. Intended to create an Application abstraction.
3. Holos prioritizes composition over abstraction.
4. An abstraction of an Application acts as a filter that removes all but the lowest common denominator functionality. The Holos Authors have found this filtering effect to create excessive friction for software developers.
5. Holos focuses instead on composition to empower developers and platform engineers to leverage the unique features and functionality of their software and platform.
## Pulumi
TODO
## Jsonnet
TODO

View File

@@ -324,28 +324,28 @@ ArgoCD, or Flux.
```mermaid
---
title: Figure 2 - Render Pipeline
title: Figure 1 - Render Pipeline
---
graph LR
PS[<a href="/docs/api/core/v1alpha2#PlatformSpec">PlatformSpec</a>]
BP[<a href="/docs/api/core/v1alpha2#BuildPlan">BuildPlan</a>]
HC[<a href="/docs/api/core/v1alpha2#HolosComponent">Components</a>]
PS[<a href="/docs/api/author/v1alpha3/#Platform">Platform</a>]
HC[<a href="/docs/api/author/v1alpha3/#ComponentFields">Components</a>]
BP[<a href="/docs/api/core/v1alpha3#BuildPlan">BuildPlan</a>]
H[<a href="/docs/api/core/v1alpha2#HelmChart">HelmChart</a>]
K[<a href="/docs/api/core/v1alpha2#KustomizeBuild">KustomizeBuild</a>]
O[<a href="/docs/api/core/v1alpha2#KubernetesObjects">KubernetesObjects</a>]
H[<a href="/docs/api/author/v1alpha3/#Helm">Helm</a>]
K[<a href="/docs/api/author/v1alpha3/#Kustomize">Kustomize</a>]
O[<a href="/docs/api/author/v1alpha3/#Kubernetes">Kubernetes</a>]
P[<a href="/docs/api/core/v1alpha2#Kustomize">Kustomize</a>]
P[<a href="/docs/api/core/v1alpha3#Kustomize">Kustomize</a>]
Y[Kubernetes <br/>Resources]
G[GitOps <br/>Resource]
FS[Local Files]
C[Kube API Server]
PS --> BP --> HC
HC --> H --> P
HC --> K --> P
HC --> O --> P
PS --> HC --> BP
BP --> H --> P
BP --> K --> P
BP --> O --> P
P --> Y --> FS
P --> G --> FS

23
doc/md/start/support.md Normal file
View File

@@ -0,0 +1,23 @@
---
description: Get Support for Holos
slug: /support
sidebar_position: 900
---
# Support
## Community Support
You can ask questions in our community forums in [GitHub Discussions](https://github.com/holos-run/holos/discussions) or [Google Groups](https://groups.google.com/g/holos-discuss).
## Commercial Support and Services
### Open Infrastructure Services
[Open Infrastructure Services] are the primary stewards of Holos. Contact Open
Infrastructure Services for training, support, and services related to Holos,
platform engineering, and cloud infrastructure automation.
Please email holos-support@openinfrastructure.co for more information.
[Open Infrastructure Services]: https://openinfrastructure.co/

View File

@@ -0,0 +1,650 @@
---
slug: technical-overview
title: Technical Overview
---
import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
import Admonition from '@theme/Admonition';
## Overview
Holos makes it easier for platform teams to integrate software into their
platform. Existing tools in the Kubernetes ecosystem are narrowly focused on
application management. Holos takes a holistic approach, focusing on the broad
integration layer where applications are joined into the platform. Holos
improves cross team collaboration through well defined, typed structures at the
integration layer. These definitions provide golden paths for other teams to
easily integrate their own services into the platform.
<!-- truncate -->
## The Problem
Platform teams need to develop and maintain significant glue code to integrate
Helm charts and YAML manifests into a platform built on Kubernetes. This glue
code is often implemented with home grown umbrella charts and scripts.
Maintaining these charts and scripts takes time and effort that could otherwise
be spent improving the platform. The need for each organization to develop and
maintain this glue code indicates a gap in the Kubernetes ecosystem. Holos is a
Go command line tool leveraging [CUE] to fill this gap.
## Key Features
1. Holos enables teams to provide simple definitions for other teams to use as golden paths.
2. Define integrations in [CUE] with strong type checking. No more text templates or bash scripts.
3. Simplify complex integration. Order does not matter. Validation is early and quick.
4. Reuse your existing Helm charts and Kustomize bases.
5. Implement the [rendered manifests pattern]. Changes are clearly visible platform-wide.
6. Fully render manifests to plain files. Use your existing GitOps tools and processes.
7. Post-process with Kustomize from CUE instead of plain text files. Customize your Kustomizations.
8. Mix in resources to Helm charts and Kustomize bases, for example ExternalSecrets.
9. Render all of Helm, Kustomize, CUE, JSON, and YAML consistently with the same process.
## Rendering Pipeline
```mermaid
---
title: Figure 1 - Render Pipeline
---
graph LR
PS[<a href="/docs/api/author/v1alpha3/#Platform">Platform</a>]
HC[<a href="/docs/api/author/v1alpha3/#ComponentFields">Components</a>]
BP[<a href="/docs/api/core/v1alpha3#BuildPlan">BuildPlan</a>]
H[<a href="/docs/api/author/v1alpha3/#Helm">Helm</a>]
K[<a href="/docs/api/author/v1alpha3/#Kustomize">Kustomize</a>]
O[<a href="/docs/api/author/v1alpha3/#Kubernetes">Kubernetes</a>]
P[<a href="/docs/api/core/v1alpha3#Kustomize">Kustomize</a>]
Y[Kubernetes <br/>Resources]
G[GitOps <br/>Resource]
FS[Local Files]
C[Kube API Server]
PS --> HC --> BP
BP --> H --> P
BP --> K --> P
BP --> O --> P
P --> Y --> FS
P --> G --> FS
FS --> ArgoCD --> C
FS --> Flux --> C
FS --> kubectl --> C
```
## Use Case
One of the development teams at the fictional Bank of Holos wants to deploy a
simple web app for an experimental project they're working on.
The platform team at the bank wants to build a simple golden path for teams to
provision projects consistently and easily in compliance with the bank's
policies.
### Platform Team
The platform team builds a golden path for development teams to register their
project with the platform. In compliance with bank policy, the platform team
needs to manage important security resources for each new project. All of these
resources can be derived from only 3 pieces of information.
1. The name of the project the dev team is working on.
2. The name of the team who currently owns the project.
3. The services, if any, the project is exposing.
The platform team defines a structure for the dev team to register this
information. This structure provides the golden path for the dev team.
The development team registers their experimental project, creatively named
"experiment" by submitting a pull request that contains this information.
<Tabs groupId="EB9C9AF1-F1AA-4189-B746-A5B8E3043F87">
<TabItem value="projects/experiment.cue" label="projects/experiment.cue">
```cue showLineNumbers
package holos
// The development team registers a project name.
#Projects: experiment: {
// The project owner must be named.
Owner: Name: "dev-team"
// Expose Service podinfo at https://podinfo.example.com
Hostnames: podinfo: Port: 9898
}
```
</TabItem>
</Tabs>
The platform team uses these three pieces of information to derive all of the
platform resources necessary to support the development team.
1. **Namespace** for the project resources.
2. **RoleBinding** to grant the dev team access to the project namespace.
3. **SecretStore** which implements the secret management policy for the bank.
4. **ReferenceGrant** to expose the project services through the Gateway API.
5. **HTTPRoutes** to expose the project services, if any.
6. **AppProject** to deploy and manage the project Applications with ArgoCD.
7. **Common Labels** to ensure every resource is labeled for resource accounting.
Rendering the platform generates fully rendered manifests for all of these
resources. These manifests are derived from the three pieces of information the
dev team provided.
Note the platform team must manage these resources across multiple namespaces.
The first four reside in the project namespace owned by the dev team. The
HTTPRoute and AppProject go into two namespaces managed by the platform team.
Holos makes it easier for the platform team to organize these resources into
different components with different owners.
:::tip
Holos supports [CODEOWNERS] by clearly defining the teams responsible for each
platform component.
:::
<Tabs groupId="2E46EA1C-B118-44BF-AE20-752E8D1CE131">
<TabItem value="command" label="Command">
```bash
holos render platform ./platform
```
</TabItem>
<TabItem value="output" label="Output">
```txt
rendered namespaces for cluster overview in 93.024042ms
rendered projects for cluster overview in 96.080667ms
rendered httproutes for cluster overview in 96.047ms
rendered platform in 96.805292ms
```
:::note
If you'd like to try this for yourself, `cd` into [examples/tech-overview] and
render the platform.
:::
</TabItem>
</Tabs>
The fully rendered manifests are written into the `deploy/` directory organized
by cluster and component for GitOps.
<Tabs groupId="07FBE14E-E9EA-437B-9FA1-C6D8806524AD">
<TabItem value="deploy/clusters/overview/components/namespaces/namespaces.gen.yaml" label="namespaces">
```yaml showLineNumbers
# deploy/clusters/overview/components/namespaces/namespaces.gen.yaml
---
metadata:
name: experiment
labels:
kubernetes.io/metadata.name: experiment
example.com/project.name: experiment
example.com/owner.name: dev-team
example.com/owner.email: sg-dev-team@example.com
kind: Namespace
apiVersion: v1
```
</TabItem>
<TabItem value="deploy/clusters/overview/components/projects/projects.gen.yaml" label="projects">
```yaml showLineNumbers
# deploy/clusters/overview/components/projects/projects.gen.yaml
---
apiVersion: gateway.networking.k8s.io/v1beta1
kind: ReferenceGrant
metadata:
name: istio-ingress
namespace: experiment
labels:
example.com/project.name: experiment
example.com/owner.name: dev-team
example.com/owner.email: sg-dev-team@example.com
spec:
from:
- group: gateway.networking.k8s.io
kind: HTTPRoute
namespace: istio-ingress
to:
- group: ""
kind: Service
---
metadata:
name: admin
namespace: experiment
labels:
example.com/project.name: experiment
example.com/owner.name: dev-team
example.com/owner.email: sg-dev-team@example.com
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: admin
subjects:
- apiGroup: rbac.authorization.k8s.io
kind: Group
name: oidc:sg-dev-team@example.com
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
---
apiVersion: external-secrets.io/v1beta1
kind: SecretStore
metadata:
name: default
namespace: experiment
labels:
example.com/project.name: experiment
example.com/owner.name: dev-team
example.com/owner.email: sg-dev-team@example.com
spec:
provider:
kubernetes:
remoteNamespace: experiment
auth:
token:
bearerToken:
key: token
name: eso-reader
server:
url: https://management.example.com:6443
caBundle: LS0tLS1CRUd...S0tLS0K
```
</TabItem>
<TabItem value="deploy/clusters/overview/components/httproutes/httproutes.gen.yaml" label="httproutes">
```yaml showLineNumbers
# deploy/clusters/overview/components/httproutes/httproutes.gen.yaml
---
apiVersion: gateway.networking.k8s.io/v1
kind: HTTPRoute
metadata:
name: podinfo.holos.localhost
namespace: istio-ingress
labels:
example.com/project.name: experiment
example.com/owner.name: dev-team
example.com/owner.email: sg-dev-team@example.com
spec:
hostnames:
- podinfo.holos.localhost
parentRefs:
- name: default
namespace: istio-ingress
rules:
- matches:
- path:
type: PathPrefix
value: /
backendRefs:
- name: podinfo
namespace: experiment
port: 9898
```
</TabItem>
</Tabs>
The rendered manifests are derived from the project registration information by
definitions implemented by the platform team. The [Author API] provides a
[Project] schema, but does not define an implementation. The platform team
implements the [Project] schema by writing a `#Projects` definition to manage
resources according to bank policies.
:::important
The Author API is intended as a convenient, ergonomic reference for component
authors. Definitions **are not** confined to the Author API.
:::
The following example shows how the platform team wrote the `#Projects`
definition to derive the Namespace from the project registration provided by the
dev team.
<Tabs groupId="5732727B-295E-46E1-B851-F8A1C5D7DF88">
<TabItem value="projects/platform/components/namespaces/namespaces.cue" label="Namespaces Component">
```txt
projects/platform/components/namespaces/namespaces.cue
```
```cue showLineNumbers
package holos
let Objects = {
Name: "namespaces"
Resources: Namespace: #Namespaces
}
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
```
1. This is the namespaces component which simply manages all of the namespaces derived from the project registration data shown in the second tab.
2. Line 5 manages a Namespace for each value of the `#Namespaces` struct. See the second tab for how the platform team defines this structure.
</TabItem>
<TabItem value="projects/projects.cue" label="#Projects Definition">
```txt
projects/projects.cue
```
```cue showLineNumbers
package holos
import api "github.com/holos-run/holos/api/author/v1alpha3"
// Projects defines the structure other teams register with to manage project
// resources. The platform team defines the schema, development teams provide
// the values.
#Projects: api.#Projects & {
[NAME=string]: {
Name: NAME
// The platform team requires the development teams to indicate an owner of
// the project.
Owner: Name: string
// The default value for the owner email address is derived from the owner
// name, but development teams can provide a different email address if
// needed.
Owner: Email: string | *"sg-\(Owner.Name)@\(#Organization.Domain)"
// The platform team constrains the project to a single namespace.
Namespaces: close({(NAME): Name: NAME})
// The platform team constrains the exposed services to the project
// namespace.
Hostnames: [HOST=string]: {
Name: HOST
Namespace: Namespaces[NAME].Name
Service: HOST
Port: number | *80
}
// CommonLabels is not part of the Projects API, so we use a hidden field to
// provide common labels to components that render resources from CUE.
_CommonLabels: {
"\(#Organization.Domain)/project.name": Name
"\(#Organization.Domain)/owner.name": Owner.Name
"\(#Organization.Domain)/owner.email": Owner.Email
}
}
}
for Project in #Projects {
// Register project namespaces with the namespaces component.
#Namespaces: {
for Namespace in Project.Namespaces {
(Namespace.Name): metadata: labels: Project._CommonLabels
}
}
}
```
1. On lines 8-37 the platform team derives most fields from the project name (line 9), and the owner name (line 13). The purpose is to fill in the remaining fields defined by the Author API.
2. Line 13 The dev team is expected to provide a concrete owner name, indicated by the `string` value.
3. Line 17 The platform team provides a default value for the email address. The project team may define a different value.
4. Line 19 The Author API allows a project to have many namespaces. The platform team constrains this down to one namespace per project by closing the struct. The namespace name must be the same as the project name.
5. Lines 22-27 The platform team derives values for a Gateway API [BackendObjectReference] from the hostname provided by the project team. These values are used later to build HTTPRoutes to expose their service.
6. Lines 31-35 Common labels aren't part of the Author API, so the platform team defines a hidden field to make them available throughout the configuration.
7. Lines 39-46 The platform team adds a namespace with common labels for each project to the struct we saw in the first tab.
</TabItem>
</Tabs>
The RoleBinding, SecretScore, and ReferenceGrant are managed in the
[projects](https://github.com/holos-run/bank-of-holos/blob/v0.1.1/examples/tech-overview/projects/platform/components/projects/projects.cue)
component, similar to the previous namespaces example.
The HTTPRoute is managed separately in the
[httproutes](https://github.com/holos-run/bank-of-holos/blob/v0.1.1/examples/tech-overview/projects/platform/components/httproutes/httproutes.cue)
component.
All components are registered with the platform in the
[platform](https://github.com/holos-run/bank-of-holos/tree/v0.1.1/examples/tech-overview/platform)
directory.
:::important
Multiple components, potentially owned by different teams, derive fully rendered
resources from the same three project values. The dev team added these three
values to the `#Projects` definition. The platform team wrote the definition to
integrate software according to bank policies. CUE powers this _unified_
platform configuration model.
:::
:::tip
Components map 1:1 to ArgoCD Applications or Flux Kustomizations.
:::
### Development Team
The development team has the platform resources they need, but they still need
to deploy their container. The development team submits a pull request adding
the following two files to deploy their existing Helm chart.
<Tabs groupId="7AD1DDA9-8001-462B-8BE0-D9410EB51233">
<TabItem value="projects/experiment/components/podinfo/podinfo.cue" label="Helm Component">
```txt
projects/experiment/components/podinfo/podinfo.cue
```
```cue showLineNumbers
package holos
// Produce a helm chart build plan.
(#Helm & Chart).BuildPlan
let Chart = {
Name: "podinfo"
Version: "6.6.2"
Repo: name: "podinfo"
Repo: url: "https://stefanprodan.github.io/podinfo"
}
```
This file represents a Helm chart component to add to the platform. The second
tab registers this component with the platform.
</TabItem>
<TabItem value="platform/podinfo.cue" label="Component Registration">
```
platform/podinfo.cue
```
```cue showLineNumbers
package holos
// Manage the component on every workload Cluster, but not management clusters.
for Cluster in #Fleets.workload.clusters {
#Platform: Components: "\(Cluster.name)/podinfo": {
path: "projects/experiment/components/podinfo"
cluster: Cluster.name
}
}
```
This file registers the component with the platform. When the platform is
rendered the dev team's Helm chart will be rendered on all workload clusters
across the platform.
</TabItem>
</Tabs>
Once the dev team's component is registered, rendering the platform will render
their component.
<Tabs groupId="1BAF7AD2-BBCD-4797-A3A6-55A626732845">
<TabItem value="command" label="Command">
```bash
holos render platform ./platform
```
</TabItem>
<TabItem value="output" label="Output">
```txt
rendered app-projects for cluster overview in 92.087042ms
rendered projects for cluster overview in 95.6325ms
rendered httproutes for cluster overview in 96.968916ms
rendered namespaces for cluster overview in 97.610291ms
// highlight-next-line
rendered podinfo for cluster overview in 155.410417ms
rendered platform in 155.470542ms
```
</TabItem>
</Tabs>
<Tabs groupId="77BF500B-105A-4AB4-A615-DEC19F501AE1">
<TabItem value="command" label="Command">
```bash
cat deploy/clusters/overview/components/podinfo/podinfo.gen.yaml
```
</TabItem>
<TabItem value="output" label="Output">
```yaml showLineNumbers
apiVersion: v1
kind: Service
metadata:
labels:
app.kubernetes.io/managed-by: Helm
app.kubernetes.io/name: podinfo
app.kubernetes.io/version: 6.6.2
example.com/owner.email: sg-dev-team@example.com
example.com/owner.name: dev-team
example.com/project.name: experiment
helm.sh/chart: podinfo-6.6.2
name: podinfo
namespace: experiment
spec:
ports:
- name: http
port: 9898
protocol: TCP
targetPort: http
- name: grpc
port: 9999
protocol: TCP
targetPort: grpc
selector:
app.kubernetes.io/name: podinfo
example.com/owner.email: sg-dev-team@example.com
example.com/owner.name: dev-team
example.com/project.name: experiment
type: ClusterIP
---
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app.kubernetes.io/managed-by: Helm
app.kubernetes.io/name: podinfo
app.kubernetes.io/version: 6.6.2
example.com/owner.email: sg-dev-team@example.com
example.com/owner.name: dev-team
example.com/project.name: experiment
helm.sh/chart: podinfo-6.6.2
name: podinfo
namespace: experiment
spec:
replicas: 1
selector:
matchLabels:
app.kubernetes.io/name: podinfo
example.com/owner.email: sg-dev-team@example.com
example.com/owner.name: dev-team
example.com/project.name: experiment
strategy:
rollingUpdate:
maxUnavailable: 1
type: RollingUpdate
template:
metadata:
annotations:
prometheus.io/port: "9898"
prometheus.io/scrape: "true"
labels:
app.kubernetes.io/name: podinfo
example.com/owner.email: sg-dev-team@example.com
example.com/owner.name: dev-team
example.com/project.name: experiment
spec:
containers:
- command:
- ./podinfo
- --port=9898
- --cert-path=/data/cert
- --port-metrics=9797
- --grpc-port=9999
- --grpc-service-name=podinfo
- --level=info
- --random-delay=false
- --random-error=false
env:
- name: PODINFO_UI_COLOR
value: '#34577c'
image: ghcr.io/stefanprodan/podinfo:6.6.2
imagePullPolicy: IfNotPresent
livenessProbe:
exec:
command:
- podcli
- check
- http
- localhost:9898/healthz
failureThreshold: 3
initialDelaySeconds: 1
periodSeconds: 10
successThreshold: 1
timeoutSeconds: 5
name: podinfo
ports:
- containerPort: 9898
name: http
protocol: TCP
- containerPort: 9797
name: http-metrics
protocol: TCP
- containerPort: 9999
name: grpc
protocol: TCP
readinessProbe:
exec:
command:
- podcli
- check
- http
- localhost:9898/readyz
failureThreshold: 3
initialDelaySeconds: 1
periodSeconds: 10
successThreshold: 1
timeoutSeconds: 5
resources:
limits: null
requests:
cpu: 1m
memory: 16Mi
volumeMounts:
- mountPath: /data
name: data
terminationGracePeriodSeconds: 30
volumes:
- emptyDir: {}
name: data
```
</TabItem>
</Tabs>
Note the rendered Helm chart resources have consistent project labels. The
platform team added a constraint to the project so all Helm charts are post
processed with Kustomize to add these common labels. The platform team
accomplishes this by adding a constraint in the project directory. This can be
seen in
[experiment/components.cue](https://github.com/holos-run/bank-of-holos/blob/v0.1.1/examples/tech-overview/projects/experiment/components.cue)
We've covered how the platform team provides a golden path for development teams
to register their projects by defining a Projects structure. We've also covered
how the development team deploys their existing Helm chart onto the platform.
## Support & Resources
1. See our [Quickstart] guide to get started with Holos.
2. Check out our other [Guides] which cover specific topics.
3. Refer to the [Author API] when writing components.
4. Consider the [Core API] if you need to do something more advanced than the Author API supports.
5. Community and commercial [Support] is available.
6. [Discussions Forum](https://github.com/holos-run/holos/discussions)
[Support]: /docs/support/
[Guides]: /docs/guides/
[API Reference]: /docs/api/
[Quickstart]: /docs/quickstart/
[CUE]: https://cuelang.org/
[Author API]: /docs/api/author/
[Core API]: /docs/api/core/
[Open Infrastructure Services]: https://openinfrastructure.co/
[Why are we templating YAML]: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&query=https%3A%2F%2Fleebriggs.co.uk%2Fblog%2F2019%2F02%2F07%2Fwhy-are-we-templating-yaml&sort=byDate&type=story
[Holos]: https://holos.run/
[Quickstart]: /docs/quickstart/
[rendered manifests pattern]: https://akuity.io/blog/the-rendered-manifests-pattern/
[examples/tech-overview]: https://github.com/holos-run/bank-of-holos/tree/v0.1.1/examples/tech-overview
[BackendObjectReference]: https://gateway-api.sigs.k8s.io/reference/spec/#gateway.networking.k8s.io%2fv1.BackendObjectReference
[CODEOWNERS]: https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/about-code-owners
[Project]: /docs/api/author/v1alpha3/#Project

View File

@@ -1,8 +0,0 @@
---
slug: welcome
title: Welcome
authors: [jeff]
tags: [holos]
---
TODO - Coming Soon

View File

@@ -0,0 +1,94 @@
---
slug: holos-platform-manager
title: Holos Platform Manager
authors: [jeff]
tags: [holos]
---
## Introducing Holos
Im excited to announce Holos, a tool designed to help engineering teams
manage their software development platforms built on the Kubernetes resource
model.
:::tip
For a hands-on introduction, check out our [Quickstart] Guide.
:::
<!-- truncate -->
### The Backstory
In our roles at [Open Infrastructure Services], and earlier at Puppet, we helped
many companies automate infrastructure management. In 2017, we had the
opportunity to work with Twitter to improve their configuration management
system. This opportunity gave us insight into the challenges of managing a
large-scale platform with multiple engineering teams. Our work involved
everything from observability systems to application deployment workflows and of
course, managing the core infrastructure.
This experience demonstrated the value of platform engineering. As the pandemic
hit, I began thinking about what a fully cloud-native platform might look like
using the Kubernetes resource model. Around the same time, I came across the
Hacker News post, “[Why Are We Templating YAML]?”, which sparked a good
discussion. It was clear I wasnt alone in my frustration with managing YAML
files and ensuring clear, predictable changes before merging them into
production.
A common pain point and theme is the complexity of working with nested YAML
configurations, especially with tools like ArgoCD and Helm. The lack of a
standard for rendering YAML templates makes it difficult to see what changes are
actually being applied to the Kubernetes API. This often results in trial and
error, costly blue-green deployments, and hours of debugging.
During the pandemic, I began experimenting with a tool to address this issue,
drawing on lessons from our work at Twitter. The key problems we aimed to solve
are:
- **Lack of visibility**: Engineers struggled to foresee the impact of small changes.
- **Large blast radius**: Small changes affected global systems, with no way to limit the impact.
- **Incomplete tooling**: While processes were in place, the right information wasnt surfaced at the right time.
We built several iterations of a reference platform based on Kubernetes,
initially focusing on fully rendering manifests into plain files—a pattern now
called the [rendered manifests pattern]. Over time, we realized we were spending
most of our time maintaining bash scripts and YAML templates. This led back to
the question: Why are we templating YAML? What _should_ replace templates?
We'd previously seen a colleague use CUE effectively to generate large scale
configurations for Envoy, and ran into CUE again when we worked on a project
involving Dagger, but I still hadn't taken a deep look at CUE.
At the end of 2023, I decided to dive deep with [CUE]. I quickly came to
appreciate CUEs unified approach where **order is irrelevant**. Before CUE, we
handled configuration data in a hierarchy with a precedence ordering, similar to
how we handled data in Puppet with Hiera. CUE's promise of no longer needing to
think about ordering and precedence rules held, alleviating a large cognitive
burden when dealing with complex configurations. CUE quickly allowed me to
replace the unmaintainable bash scripts and complex Helm templates, simplifying
our workflow.
### Enter Holos
Holos adds CUE as a well-specified integration layer over tools like Helm,
Kustomize, ArgoCD, and Crossplane. With Holos, we can now efficiently integrate
upstream Helm charts and Kustomize bases into our platform without the
complexity of templates and scripts. This has also made it easy for one team to
define "golden paths" that other teams can follow—like automatically configuring
namespaces and security policies when dev teams start new projects.
We've found Holos incredibly useful and hope you do too. Let us know your
thoughts!
[Guides]: /docs/guides/
[API Reference]: /docs/api/
[Quickstart]: /docs/quickstart/
[CUE]: https://cuelang.org/
[Author API]: /docs/api/author/
[Core API]: /docs/api/core/
[Open Infrastructure Services]: https://openinfrastructure.co/
[Why are we templating YAML]: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&query=https%3A%2F%2Fleebriggs.co.uk%2Fblog%2F2019%2F02%2F07%2Fwhy-are-we-templating-yaml&sort=byDate&type=story
[Holos]: https://holos.run/
[Quickstart]: /docs/quickstart/
[rendered manifests pattern]: https://akuity.io/blog/the-rendered-manifests-pattern/

View File

@@ -1,6 +1,6 @@
jeff:
name: Jeff McCune
title: Holos maintainer & creator
title: Holos maintainer & author
url: https://github.com/jeffmccune
image_url: https://github.com/jeffmccune.png

View File

@@ -2,3 +2,8 @@ holos:
label: Holos
permalink: /holos
description: Holos Platform
tech:
label: Technical
permalink: /tech
description: Technical Articles

View File

@@ -4,7 +4,7 @@ import type * as Preset from '@docusaurus/preset-classic';
const config: Config = {
title: 'Holos',
tagline: 'The Holistic Package Manager for Cloud Native Applications',
tagline: 'An easier way for platform teams to integrate software into their platform',
favicon: 'img/favicon.ico',
// Set the production url of your site here
@@ -12,6 +12,7 @@ const config: Config = {
// Set the /<baseUrl>/ pathname under which your site is served
// For GitHub pages deployment, it is often '/<projectName>/'
baseUrl: '/',
// trailing slash is necessary for Cloudflare pages.
trailingSlash: true,
// GitHub pages deployment config.
@@ -65,13 +66,17 @@ const config: Config = {
blogSidebarTitle: "All posts",
feedOptions: {
type: 'all',
copyright: `Copyright © ${new Date().getFullYear()}, The Holos Authors.`,
copyright: `Copyright © ${new Date().getFullYear()}, The Holos Authors`,
},
showReadingTime: false,
},
theme: {
customCss: './src/css/custom.css',
},
gtag: {
trackingID: 'G-M00QMB1N05',
anonymizeIP: true,
}
} satisfies Preset.Options,
],
],
@@ -95,9 +100,10 @@ const config: Config = {
type: 'doc',
docId: 'guides/quickstart',
position: 'left',
label: 'Try Holos',
label: 'Quickstart',
},
{ to: '/docs', label: 'Docs', position: 'left' },
{ to: '/docs/technical-overview', label: 'Docs', position: 'left' },
{ to: '/docs/guides', label: 'Guides', position: 'left' },
{
type: 'doc',
docId: 'api',
@@ -106,13 +112,7 @@ const config: Config = {
},
{ to: '/blog', label: 'Blog', position: 'left' },
{
"href": "https://pkg.go.dev/github.com/holos-run/holos?tab=doc",
"label": "GoDoc",
"position": "left",
"className": "header-godoc-link",
},
{
href: 'https://github.com/holos-run/holos',
href: 'https://github.com/holos-run',
label: 'GitHub',
position: 'right',
},
@@ -146,7 +146,19 @@ const config: Config = {
title: 'Community',
items: [
{
label: 'Discuss',
label: 'Support',
href: '/docs/support',
},
{
label: 'Announcements List',
href: 'https://groups.google.com/g/holos-announce',
},
{
label: 'Discussion List',
href: 'https://groups.google.com/g/holos-discuss',
},
{
label: 'Discussion Forum',
href: 'https://github.com/holos-run/holos/discussions',
},
],
@@ -162,6 +174,10 @@ const config: Config = {
label: 'GitHub',
href: 'https://github.com/holos-run/holos',
},
{
label: 'GoDoc',
href: 'https://pkg.go.dev/github.com/holos-run/holos?tab=doc',
}
],
},
],

View File

@@ -13,6 +13,7 @@ import type { SidebarsConfig } from '@docusaurus/plugin-content-docs';
const sidebars: SidebarsConfig = {
doc: [
'introduction',
'technical-overview',
{
label: 'Getting Started',
type: 'category',
@@ -44,14 +45,14 @@ const sidebars: SidebarsConfig = {
link: { type: 'doc', id: 'api' },
items: [
{
label: 'Schema API',
label: 'Author API',
type: 'category',
link: { type: 'doc', id: 'api/schema' },
link: { type: 'doc', id: 'api/author' },
collapsed: true,
items: [
{
type: 'autogenerated',
dirName: 'api/schema',
dirName: 'api/author',
},
]
},

View File

@@ -8,53 +8,56 @@ type FeatureItem = {
description: JSX.Element;
};
// TODO: Consider focusing on the three pillars of Safe, Easy, Consistent.
// We don't focus on features, but rather problems and solutions.
const FeatureList: FeatureItem[] = [
{
title: 'Kustomize Helm',
Svg: require('@site/static/img/base00/undraw_together_re_a8x4.svg').default,
title: 'For Platform Engineers',
Svg: require('@site/static/img/base00/undraw_software_engineer_re_tnjc.svg').default,
description: (
<>
Super charge your existing Helm charts by providing well defined,
validated input values, post-processing the output with Kustomize,
and mixing in your own custom resources. All without modifying upstream
charts to alleviate the pain of upgrades.
<p align="left">
<ul>
<li>Provide simple definitions for other teams to use as golden paths.</li>
<li>Define integrations in <a href="https://cuelang.org/">CUE</a> with strong type checking. No more text templates or bash scripts.</li>
<li>Reuse your existing Helm charts and Kustomize bases.</li>
</ul>
</p>
<a href="/docs/technical-overview">Learn More</a>
</>
),
},
{
title: 'Unified Data Model',
Svg: require('@site/static/img/base00/undraw_fitting_pieces_re_nss7.svg').default,
title: 'For Software Developers',
Svg: require('@site/static/img/base00/undraw_through_the_park_lxnl.svg').default,
description: (
<>
Unify all of your platform components into one well-defined, strongly
typed data model with CUE. Holos makes it easier and safer to integrate
seamlessly with software distributed with current and future tools that
produce Kubernetes resource manifests.
<p align="left">
<ul>
<li>Move faster using paved paths from your platform and security teams.</li>
<li>Develop locally or in the cloud.</li>
<li>Spend more time developing software and fewer cycles fighting infrastructure challenges.</li>
</ul>
</p>
<a href="/docs/technical-overview">Learn More</a>
</>
),
},
{
title: 'Deep Insights',
Svg: require('@site/static/img/base00/undraw_code_review_re_woeb.svg').default,
title: 'For Security Teams',
Svg: require('@site/static/img/base00/undraw_security_on_re_e491.svg').default,
description: (
<>
Reduce risk and increase confidence in your configuration changes.
Holos offers clear visibility into complete resource configuration
<i>before</i> being applied.
<p align="left">
<ul>
<li>Define security policy as reusable, typed configurations.</li>
<li>Automatically enforce security policy on new projects.</li>
<li>Ensure a consistent security posture cross-platform with fewer code changes.</li>
</ul>
</p>
<a href="/docs/technical-overview">Learn More</a>
</>
),
},
{
title: 'Interoperable',
Svg: require('@site/static/img/base00/undraw_version_control_re_mg66.svg').default,
description: (
<>
Holos is designed for compatibility with your preferred tools and
processes, for example git diff and code reviews.
</>
),
},
}
];
function Feature({ title, Svg, description }: FeatureItem) {

View File

@@ -5,9 +5,40 @@
*/
/* Enable wrapping by default for mobile */
pre code {
/* pre code {
white-space: pre-wrap;
overflow-wrap: anywhere;
} */
.hero__title {
text-align: left;
}
.hero__subtitle {
text-align: left;
}
.projectDesc {
text-align: left;
}
.hero__buttons {
float: left
}
/* Ensure img in hero banner scales well even on mobile */
@media screen and (max-width: 996px) {
div.diagramImg {
width: 100%;
max-width: 100px;
height: auto;
}
}
div.diagramImg {
width: 30%;
min-width: 300px;
float: right;
}
/* You can override the default Infima variables here. */

View File

@@ -12,30 +12,30 @@ function HomepageHeader() {
return (
<header className={clsx('hero hero--primary', styles.heroBanner)}>
<div className="container">
<div className="diagramImg">
<img src="./img/holos-diagram-color-transparent.svg" alt="Holos Diagram" />
</div>
<Heading as="h1" className="hero__title">
{siteConfig.title}
</Heading>
<p className="hero__subtitle">{siteConfig.tagline}</p>
<p className="projectDesc">
Holos adds CUE's type safety, unified structure, and strong validation
features to your current software packages, including Helm and
Kustomize. These features make the experience of integrating software
packages into a holistic platform a pleasant journey.
</p>
<div className={styles.buttons}>
<Link
className="button button--secondary button--lg"
to="docs/quickstart">
Get Started
</Link>
<span className={styles.divider}></span>
<Link
className="button button--primary button--lg"
to="docs/concepts">
Learn More
</Link>
<div className="hero__buttons">
<div className={styles.buttons}>
<Link
className="button button--secondary button--lg"
to="docs/quickstart">
Get Started
</Link>
<span className={styles.divider}></span>
<Link
className="button button--primary button--lg"
to="docs/technical-overview/">
Learn More
</Link>
<span className={styles.divider}></span>
</div>
</div>
</div>
</div >
</header >
);
}
@@ -44,8 +44,8 @@ export default function Home(): JSX.Element {
const { siteConfig } = useDocusaurusContext();
return (
<Layout
title={`${siteConfig.title} Package Manager`}
description="Holos adds CUE's type safety, unified structure, and strong validation features to your current software packages, including Helm and Kustomize.">
title={`${siteConfig.title} Platform Manager`}
description="Holos adds CUE's type safety, unified structure, and strong validation features to your Kubernetes configuration manifests, including Helm and Kustomize.">
<HomepageHeader />
<main>
<HomepageFeatures />

File diff suppressed because one or more lines are too long

After

Width:  |  Height:  |  Size: 33 KiB

File diff suppressed because one or more lines are too long

After

Width:  |  Height:  |  Size: 33 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 461 KiB

After

Width:  |  Height:  |  Size: 480 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 477 KiB

View File

@@ -4,4 +4,4 @@ set -euo pipefail
TOPLEVEL="$(cd $(dirname "$0") && git rev-parse --show-toplevel)"
cd "${TOPLEVEL}" && npx cspell ./doc/md/**/*.{md,mdx,markdown}
cd "${TOPLEVEL}" && npx cspell ./doc/md/**/*.{md,mdx,markdown} ./doc/md/*.{md,mdx,markdown}

View File

@@ -1,6 +1,8 @@
// Package holos defines types for the rest of the system.
package holos
import "context"
// A PathCueMod is a string representing the absolute filesystem path of a cue
// module. It is given a unique type so the API is clear.
type PathCueMod string
@@ -8,3 +10,33 @@ type PathCueMod string
// A InstancePath is a string representing the absolute filesystem path of a
// holos instance. It is given a unique type so the API is clear.
type InstancePath string
// FilePath represents the path of a file relative to the current working
// directory of holos at runtime.
type FilePath string
// FileContent represents the contents of a file as a string.
type FileContent string
// TypeMeta represents the kind and version of a resource holos needs to
// process. Useful to discriminate generated resources.
type TypeMeta struct {
Kind string `json:"kind,omitempty" yaml:"kind,omitempty"`
APIVersion string `json:"apiVersion,omitempty" yaml:"apiVersion,omitempty"`
}
// Builder builds file artifacts.
type Builder interface {
Build(context.Context, ArtifactMap) error
}
// ArtifactMap sets and gets data for file artifacts.
//
// Concrete values must ensure Set is write once, returning an error if a given
// FilePath was previously Set. Concrete values must be safe for concurrent
// reads and writes.
type ArtifactMap interface {
Get(path string) (data []byte, ok bool)
Set(path string, data []byte) error
Save(dir, path string) error
}

View File

@@ -0,0 +1,69 @@
package artifact
import (
"fmt"
"os"
"path/filepath"
"sync"
"github.com/holos-run/holos/internal/errors"
)
func New() *Artifact {
return &Artifact{m: make(map[string][]byte)}
}
// Artifact represents the fully rendered manifests build from the holos
// rendering pipeline. Files are organized by keys representing paths relative
// to the current working directory. Values represent the file content.
type Artifact struct {
mu sync.RWMutex
m map[string][]byte
}
// Set sets an artifact file with write locking. Set returns an error if the
// artifact was previously set.
func (a *Artifact) Set(path string, data []byte) error {
a.mu.Lock()
defer a.mu.Unlock()
if _, ok := a.m[path]; ok {
return errors.Format("%s already set", path)
}
a.m[path] = data
return nil
}
// Get gets the content of an artifact with read locking.
func (a *Artifact) Get(path string) (data []byte, ok bool) {
a.mu.RLock()
defer a.mu.RUnlock()
data, ok = a.m[path]
return
}
// Save writes a file to the filesystem.
func (a *Artifact) Save(dir, path string) error {
fullPath := filepath.Join(dir, path)
msg := fmt.Sprintf("could not save %s", fullPath)
data, ok := a.Get(path)
if !ok {
return errors.Format("%s: missing %s", msg, path)
}
if err := os.MkdirAll(filepath.Dir(fullPath), 0777); err != nil {
return errors.Format("%s: %w", msg, err)
}
if err := os.WriteFile(fullPath, data, 0666); err != nil {
return errors.Format("%s: %w", msg, err)
}
return nil
}
func (a *Artifact) Keys() []string {
a.mu.RLock()
defer a.mu.RUnlock()
keys := make([]string, 0, len(a.m))
for key := range a.m {
keys = append(keys, key)
}
return keys
}

View File

@@ -15,10 +15,12 @@ import (
"cuelang.org/go/cue"
"cuelang.org/go/cue/cuecontext"
"cuelang.org/go/cue/load"
v1 "github.com/holos-run/holos/api/core/v1alpha3"
"github.com/holos-run/holos/api/v1alpha1"
"github.com/holos-run/holos"
core_v1alpha2 "github.com/holos-run/holos/api/core/v1alpha2"
coreA3 "github.com/holos-run/holos/api/core/v1alpha3"
metaA2 "github.com/holos-run/holos/api/meta/v1alpha2"
"github.com/holos-run/holos/api/v1alpha1"
"github.com/holos-run/holos/internal/client"
"github.com/holos-run/holos/internal/errors"
"github.com/holos-run/holos/internal/logger"
@@ -26,10 +28,10 @@ import (
)
const (
KubernetesObjects = v1.KubernetesObjectsKind
KubernetesObjects = coreA3.KubernetesObjectsKind
// Helm is the value of the kind field of holos build output indicating helm
// values and helm command information.
Helm = v1.HelmChartKind
Helm = coreA3.HelmChartKind
// Skip is the value when the instance should be skipped
Skip = "Skip"
// KustomizeBuild is the value of the kind field of cue output indicating
@@ -43,11 +45,7 @@ type Option func(*config)
type config struct {
args []string
cluster string
}
type Builder struct {
cfg config
ctx *cue.Context
tags []string
}
// BuildData represents the data necessary to produce a build plan. It is a
@@ -59,8 +57,26 @@ type BuildData struct {
Dir string
}
func (bd *BuildData) TypeMeta() (tm holos.TypeMeta, err error) {
jsonBytes, err := bd.Value.MarshalJSON()
if err != nil {
err = errors.Format("could not marshal json %s: %w", bd.Dir, err)
return
}
err = json.NewDecoder(bytes.NewReader(jsonBytes)).Decode(&tm)
if err != nil {
err = errors.Format("could not decode type meta %s: %w", bd.Dir, err)
}
return
}
type Builder struct {
cfg config
ctx *cue.Context
}
type buildPlanWrapper struct {
buildPlan *v1.BuildPlan
buildPlan *coreA3.BuildPlan
}
func (b *buildPlanWrapper) validate() error {
@@ -72,7 +88,7 @@ func (b *buildPlanWrapper) validate() error {
return fmt.Errorf("invalid BuildPlan: is nil")
}
errs := make([]string, 0, 2)
if bp.Kind != v1.BuildPlanKind {
if bp.Kind != coreA3.BuildPlanKind {
errs = append(errs, fmt.Sprintf("kind invalid: want: %s have: %s", v1alpha1.BuildPlanKind, bp.Kind))
}
if len(errs) > 0 {
@@ -116,6 +132,11 @@ func Cluster(name string) Option {
return func(cfg *config) { cfg.cluster = name }
}
// Tags configures tags to pass to cue when building the instance.
func Tags(tags []string) Option {
return func(cfg *config) { cfg.tags = tags }
}
// Cluster returns the cluster name of the component instance being built.
func (b *Builder) Cluster() string {
return b.cfg.cluster
@@ -141,15 +162,19 @@ func (b *Builder) Unify(ctx context.Context, cfg *client.Config) (bd BuildData,
return bd, errors.Wrap(fmt.Errorf("could not load platform model: %w", err))
}
tags := make([]string, 0, len(b.cfg.tags)+2)
tags = append(tags,
"cluster="+cfg.Holos().ClusterName(),
// TODO: Use instance.FillPath to fill the platform config.
// Refer to https://pkg.go.dev/cuelang.org/go/cue#Value.FillPath
"platform_config="+string(platformConfigData),
)
tags = append(tags, b.cfg.tags...)
cueConfig := load.Config{
Dir: bd.ModuleRoot,
ModuleRoot: bd.ModuleRoot,
// TODO: Use instance.FillPath to fill the platform config.
// Refer to https://pkg.go.dev/cuelang.org/go/cue#Value.FillPath
Tags: []string{
"cluster=" + cfg.Holos().ClusterName(),
"platform_config=" + string(platformConfigData),
},
Tags: tags,
}
// Make args relative to the module directory
@@ -229,6 +254,8 @@ func loadUserData(ctx *cue.Context, moduleRoot string) (val cue.Value, err error
// provided to the entrypoint through a json encoded string tag named
// platform_config. The resulting cue.Value is unified with all user data files
// at the path "#UserData".
//
// Deprecated: Use holos.Builder instead
func (b *Builder) Run(ctx context.Context, cfg *client.Config) (results []*render.Result, err error) {
log := logger.FromContext(ctx)
log.DebugContext(ctx, "cue: building instances")
@@ -240,7 +267,7 @@ func (b *Builder) Run(ctx context.Context, cfg *client.Config) (results []*rende
return b.build(ctx, bd)
}
func (b Builder) build(ctx context.Context, bd BuildData) (results []*render.Result, err error) {
func (b *Builder) build(ctx context.Context, bd BuildData) (results []*render.Result, err error) {
log := logger.FromContext(ctx).With("dir", bd.InstancePath)
value := bd.Value
@@ -279,7 +306,7 @@ func (b Builder) build(ctx context.Context, bd BuildData) (results []*render.Res
switch tm.Kind {
case "BuildPlan":
var bp v1.BuildPlan
var bp coreA3.BuildPlan
if err = decoder.Decode(&bp); err != nil {
err = errors.Wrap(fmt.Errorf("could not decode BuildPlan %s: %w", bd.Dir, err))
return
@@ -295,7 +322,7 @@ func (b Builder) build(ctx context.Context, bd BuildData) (results []*render.Res
return results, err
}
func (b *Builder) buildPlan(ctx context.Context, buildPlan *v1.BuildPlan, path holos.InstancePath) (results []*render.Result, err error) {
func (b *Builder) buildPlan(ctx context.Context, buildPlan *coreA3.BuildPlan, path holos.InstancePath) (results []*render.Result, err error) {
log := logger.FromContext(ctx)
bpw := buildPlanWrapper{buildPlan: buildPlan}
@@ -381,3 +408,62 @@ func (b *Builder) findCueMod() (dir holos.PathCueMod, err error) {
}
return dir, nil
}
// Platform builds a platform
// TODO: Refactor, lift up into NewPlatform RunE.
func (b *Builder) Platform(ctx context.Context, cfg *client.Config) (*core_v1alpha2.Platform, error) {
log := logger.FromContext(ctx)
log.DebugContext(ctx, "cue: building platform instance")
bd, err := b.Unify(ctx, cfg)
if err != nil {
return nil, errors.Wrap(err)
}
return b.runPlatform(ctx, bd)
}
func (b *Builder) runPlatform(ctx context.Context, bd BuildData) (*core_v1alpha2.Platform, error) {
path := holos.InstancePath(bd.Dir)
log := logger.FromContext(ctx).With("dir", path)
value := bd.Value
if err := bd.Value.Err(); err != nil {
return nil, errors.Wrap(fmt.Errorf("could not load: %w", err))
}
log.DebugContext(ctx, "cue: validating instance")
if err := value.Validate(); err != nil {
return nil, errors.Wrap(fmt.Errorf("could not validate: %w", err))
}
log.DebugContext(ctx, "cue: decoding holos platform")
jsonBytes, err := value.MarshalJSON()
if err != nil {
return nil, errors.Wrap(fmt.Errorf("could not marshal cue instance %s: %w", bd.Dir, err))
}
decoder := json.NewDecoder(bytes.NewReader(jsonBytes))
// Discriminate the type of build plan.
tm := &metaA2.TypeMeta{}
err = decoder.Decode(tm)
if err != nil {
return nil, errors.Wrap(fmt.Errorf("invalid platform: %s: %w", bd.Dir, err))
}
log.DebugContext(ctx, "cue: discriminated build kind: "+tm.GetKind(), "kind", tm.GetKind(), "apiVersion", tm.GetAPIVersion())
decoder = json.NewDecoder(bytes.NewReader(jsonBytes))
decoder.DisallowUnknownFields()
var pf core_v1alpha2.Platform
switch tm.GetKind() {
case "Platform":
if err = decoder.Decode(&pf); err != nil {
err = errors.Wrap(fmt.Errorf("could not decode platform %s: %w", bd.Dir, err))
return nil, err
}
return &pf, nil
default:
err = errors.Wrap(fmt.Errorf("unknown kind: %v", tm.GetKind()))
}
return nil, err
}

View File

@@ -1,75 +0,0 @@
package builder
import (
"bytes"
"context"
"encoding/json"
"fmt"
"github.com/holos-run/holos"
core "github.com/holos-run/holos/api/core/v1alpha2"
meta "github.com/holos-run/holos/api/meta/v1alpha2"
"github.com/holos-run/holos/internal/client"
"github.com/holos-run/holos/internal/errors"
"github.com/holos-run/holos/internal/logger"
)
// Platform builds a platform
func (b *Builder) Platform(ctx context.Context, cfg *client.Config) (*core.Platform, error) {
log := logger.FromContext(ctx)
log.DebugContext(ctx, "cue: building platform instance")
bd, err := b.Unify(ctx, cfg)
if err != nil {
return nil, errors.Wrap(err)
}
return b.runPlatform(ctx, bd)
}
func (b Builder) runPlatform(ctx context.Context, bd BuildData) (*core.Platform, error) {
path := holos.InstancePath(bd.Dir)
log := logger.FromContext(ctx).With("dir", path)
value := bd.Value
if err := bd.Value.Err(); err != nil {
return nil, errors.Wrap(fmt.Errorf("could not load: %w", err))
}
log.DebugContext(ctx, "cue: validating instance")
if err := value.Validate(); err != nil {
return nil, errors.Wrap(fmt.Errorf("could not validate: %w", err))
}
log.DebugContext(ctx, "cue: decoding holos platform")
jsonBytes, err := value.MarshalJSON()
if err != nil {
return nil, errors.Wrap(fmt.Errorf("could not marshal cue instance %s: %w", bd.Dir, err))
}
decoder := json.NewDecoder(bytes.NewReader(jsonBytes))
// Discriminate the type of build plan.
tm := &meta.TypeMeta{}
err = decoder.Decode(tm)
if err != nil {
return nil, errors.Wrap(fmt.Errorf("invalid platform: %s: %w", bd.Dir, err))
}
log.DebugContext(ctx, "cue: discriminated build kind: "+tm.GetKind(), "kind", tm.GetKind(), "apiVersion", tm.GetAPIVersion())
// New decoder for the full object
decoder = json.NewDecoder(bytes.NewReader(jsonBytes))
decoder.DisallowUnknownFields()
var pf core.Platform
switch tm.GetKind() {
case "Platform":
if err = decoder.Decode(&pf); err != nil {
err = errors.Wrap(fmt.Errorf("could not decode platform %s: %w", bd.Dir, err))
return nil, err
}
return &pf, nil
default:
err = errors.Wrap(fmt.Errorf("unknown kind: %v", tm.GetKind()))
}
return nil, err
}

View File

@@ -0,0 +1,543 @@
package v1alpha4
import (
"bytes"
"context"
"fmt"
"io"
"log/slog"
"os"
"path/filepath"
"strings"
"syscall"
"time"
h "github.com/holos-run/holos"
"github.com/holos-run/holos/api/core/v1alpha4"
"github.com/holos-run/holos/internal/errors"
"github.com/holos-run/holos/internal/logger"
"github.com/holos-run/holos/internal/util"
"golang.org/x/sync/errgroup"
"gopkg.in/yaml.v3"
)
// Platform represents a platform builder.
type Platform struct {
Platform v1alpha4.Platform
Concurrency int
Stderr io.Writer
}
// Build builds a Platform by concurrently building a BuildPlan for each
// platform component. No artifact files are written directly, only indirectly
// by rendering each component.
func (p *Platform) Build(ctx context.Context, _ h.ArtifactMap) error {
parentStart := time.Now()
components := p.Platform.Spec.Components
total := len(components)
g, ctx := errgroup.WithContext(ctx)
// Limit the number of concurrent goroutines due to CUE memory usage concerns
// while rendering components. One more for the producer.
g.SetLimit(p.Concurrency + 1)
// Spawn a producer because g.Go() blocks when the group limit is reached.
g.Go(func() error {
for idx := range components {
select {
case <-ctx.Done():
return ctx.Err()
default:
// Capture idx to avoid issues with closure. Fixed in Go 1.22.
idx := idx
component := &components[idx]
// Worker go routine. Blocks if limit has been reached.
g.Go(func() error {
select {
case <-ctx.Done():
return ctx.Err()
default:
start := time.Now()
log := logger.FromContext(ctx).With(
"name", component.Name,
"path", component.Component,
"cluster", component.Cluster,
"environment", component.Environment,
"num", idx+1,
"total", total,
)
log.DebugContext(ctx, "render component")
tags := make([]string, 0, 3+len(component.Tags))
tags = append(tags, "name="+component.Name)
tags = append(tags, "component="+component.Component)
tags = append(tags, "environment="+component.Environment)
// Tags are unified, cue handles conflicts. We don't bother.
tags = append(tags, component.Tags...)
// Execute a sub-process to limit CUE memory usage.
args := []string{
"render",
"component",
"--cluster-name", component.Cluster,
"--tags", strings.Join(tags, ","),
component.Component,
}
result, err := util.RunCmd(ctx, "holos", args...)
// I've lost an hour+ digging into why I couldn't see log output
// from sub-processes. Make sure to surface at least stderr from
// sub-processes.
_, _ = io.Copy(p.Stderr, result.Stderr)
if err != nil {
return errors.Wrap(fmt.Errorf("could not render component: %w", err))
}
duration := time.Since(start)
msg := fmt.Sprintf(
"rendered %s for cluster %s in %s",
component.Name,
component.Cluster,
duration,
)
log.InfoContext(ctx, msg, "duration", duration)
return nil
}
})
}
}
return nil
})
// Wait for completion and return the first error (if any)
if err := g.Wait(); err != nil {
return err
}
duration := time.Since(parentStart)
msg := fmt.Sprintf("rendered platform in %s", duration)
logger.FromContext(ctx).InfoContext(ctx, msg, "duration", duration, "version", p.Platform.APIVersion)
return nil
}
// BuildPlan represents a component builder.
type BuildPlan struct {
BuildPlan v1alpha4.BuildPlan
Concurrency int
Stderr io.Writer
// WriteTo --write-to=deploy flag
WriteTo string
// Path represents the path to the component
Path h.InstancePath
}
// Build builds a BuildPlan into Artifact files.
func (b *BuildPlan) Build(ctx context.Context, am h.ArtifactMap) error {
name := b.BuildPlan.Metadata.Name
component := b.BuildPlan.Spec.Component
log := logger.FromContext(ctx).With("name", name, "component", component)
msg := fmt.Sprintf("could not build %s", name)
if b.BuildPlan.Spec.Disabled {
log.WarnContext(ctx, fmt.Sprintf("%s: disabled", msg))
return nil
}
g, ctx := errgroup.WithContext(ctx)
// One more for the producer
g.SetLimit(b.Concurrency + 1)
// Producer.
g.Go(func() error {
for _, a := range b.BuildPlan.Spec.Artifacts {
msg := fmt.Sprintf("%s artifact %s", msg, a.Artifact)
log := log.With("artifact", a.Artifact)
if a.Skip {
log.WarnContext(ctx, fmt.Sprintf("%s: skipped field is true", msg))
continue
}
select {
case <-ctx.Done():
return ctx.Err()
default:
// https://golang.org/doc/faq#closures_and_goroutines
a := a
// Worker. Blocks if limit has been reached.
g.Go(func() error {
for _, gen := range a.Generators {
switch gen.Kind {
case "Resources":
if err := b.resources(log, gen, am); err != nil {
return errors.Format("could not generate resources: %w", err)
}
case "Helm":
if err := b.helm(ctx, log, gen, am); err != nil {
return errors.Format("could not generate helm: %w", err)
}
case "File":
if err := b.file(log, gen, am); err != nil {
return errors.Format("could not generate file: %w", err)
}
default:
return errors.Format("%s: unsupported kind %s", msg, gen.Kind)
}
}
for _, t := range a.Transformers {
switch t.Kind {
case "Kustomize":
if err := b.kustomize(ctx, log, t, am); err != nil {
return errors.Wrap(err)
}
case "Join":
s := make([][]byte, 0, len(t.Inputs))
for _, input := range t.Inputs {
if data, ok := am.Get(string(input)); ok {
s = append(s, data)
} else {
return errors.Format("%s: missing %s", msg, input)
}
}
data := bytes.Join(s, []byte(t.Join.Separator))
if err := am.Set(string(t.Output), data); err != nil {
return errors.Format("%s: %w", msg, err)
}
log.Debug("set artifact: " + string(t.Output))
default:
return errors.Format("%s: unsupported kind %s", msg, t.Kind)
}
}
// Write the final artifact
if err := am.Save(b.WriteTo, string(a.Artifact)); err != nil {
return errors.Format("%s: %w", msg, err)
}
log.DebugContext(ctx, "wrote "+filepath.Join(b.WriteTo, string(a.Artifact)))
return nil
})
}
}
return nil
})
// Wait for completion and return the first error (if any)
return g.Wait()
}
func (b *BuildPlan) file(
log *slog.Logger,
g v1alpha4.Generator,
am h.ArtifactMap,
) error {
data, err := os.ReadFile(filepath.Join(string(b.Path), string(g.File.Source)))
if err != nil {
return errors.Wrap(err)
}
if err := am.Set(string(g.Output), data); err != nil {
return errors.Wrap(err)
}
log.Debug("set artifact: " + string(g.Output))
return nil
}
func (b *BuildPlan) helm(
ctx context.Context,
log *slog.Logger,
g v1alpha4.Generator,
am h.ArtifactMap,
) error {
chartName := g.Helm.Chart.Name
log = log.With("chart", chartName)
// Unnecessary? cargo cult copied from internal/cli/render/render.go
if chartName == "" {
return errors.New("missing chart name")
}
// Cache the chart by version to pull new versions. (#273)
cacheDir := filepath.Join(string(b.Path), "vendor", g.Helm.Chart.Version)
cachePath := filepath.Join(cacheDir, filepath.Base(chartName))
if _, err := os.Stat(cachePath); os.IsNotExist(err) {
timeout, cancel := context.WithTimeout(ctx, 5*time.Minute)
defer cancel()
err := onceWithLock(log, timeout, cachePath, func() error {
return b.cacheChart(ctx, log, cacheDir, g.Helm.Chart)
})
if err != nil {
return errors.Format("could not cache chart: %w", err)
}
}
// Write values file
tempDir, err := os.MkdirTemp("", "holos.helm")
if err != nil {
return errors.Format("could not make temp dir: %w", err)
}
defer util.Remove(ctx, tempDir)
data, err := yaml.Marshal(g.Helm.Values)
if err != nil {
return errors.Format("could not marshal values: %w", err)
}
valuesPath := filepath.Join(tempDir, "values.yaml")
if err := os.WriteFile(valuesPath, data, 0666); err != nil {
return errors.Wrap(fmt.Errorf("could not write values: %w", err))
}
log.DebugContext(ctx, "wrote"+valuesPath)
// Run charts
args := []string{"template"}
if g.Helm.EnableHooks {
args = append(args, "--hooks")
} else {
args = append(args, "--no-hooks")
}
args = append(args,
"--include-crds",
"--values", valuesPath,
"--namespace", g.Helm.Namespace,
"--kubeconfig", "/dev/null",
"--version", g.Helm.Chart.Version,
g.Helm.Chart.Release,
cachePath,
)
helmOut, err := util.RunCmd(ctx, "helm", args...)
if err != nil {
stderr := helmOut.Stderr.String()
lines := strings.Split(stderr, "\n")
for _, line := range lines {
if strings.HasPrefix(line, "Error:") {
err = fmt.Errorf("%s: %w", line, err)
}
}
return errors.Format("could not run helm template: %w", err)
}
// Set the artifact
if err := am.Set(string(g.Output), helmOut.Stdout.Bytes()); err != nil {
return errors.Format("could not store helm output: %w", err)
}
log.Debug("set artifact: " + string(g.Output))
return nil
}
func (b *BuildPlan) resources(
log *slog.Logger,
g v1alpha4.Generator,
am h.ArtifactMap,
) error {
var size int
for _, m := range g.Resources {
size += len(m)
}
list := make([]v1alpha4.Resource, 0, size)
for _, m := range g.Resources {
for _, r := range m {
list = append(list, r)
}
}
msg := fmt.Sprintf("could not generate %s for %s", g.Output, b.BuildPlan.Metadata.Name)
buf, err := marshal(list)
if err != nil {
return errors.Format("%s: %w", msg, err)
}
if err := am.Set(string(g.Output), buf.Bytes()); err != nil {
return errors.Format("%s: %w", msg, err)
}
log.Debug("set artifact " + string(g.Output))
return nil
}
func (b *BuildPlan) kustomize(
ctx context.Context,
log *slog.Logger,
t v1alpha4.Transformer,
am h.ArtifactMap,
) error {
tempDir, err := os.MkdirTemp("", "holos.kustomize")
if err != nil {
return errors.Wrap(err)
}
defer util.Remove(ctx, tempDir)
msg := fmt.Sprintf("could not transform %s for %s", t.Output, b.BuildPlan.Metadata.Name)
// Write the kustomization
data, err := yaml.Marshal(t.Kustomize.Kustomization)
if err != nil {
return errors.Format("%s: %w", msg, err)
}
path := filepath.Join(tempDir, "kustomization.yaml")
if err := os.WriteFile(path, data, 0666); err != nil {
return errors.Format("%s: %w", msg, err)
}
log.DebugContext(ctx, "wrote "+path)
// Write the inputs
for _, input := range t.Inputs {
path := string(input)
if err := am.Save(tempDir, path); err != nil {
return errors.Format("%s: %w", msg, err)
}
log.DebugContext(ctx, "wrote "+filepath.Join(tempDir, path))
}
// Execute kustomize
r, err := util.RunCmd(ctx, "kubectl", "kustomize", tempDir)
if err != nil {
err = errors.Format("%s: could not run kustomize: %w", msg, err)
if s := strings.ReplaceAll(r.Stderr.String(), "\n", "\n\t"); s != "" {
err = errors.Format("%w\n\t%s", err, s)
}
return err
}
// Store the artifact
if err := am.Set(string(t.Output), r.Stdout.Bytes()); err != nil {
return errors.Format("%s: %w", msg, err)
}
log.Debug("set artifact " + string(t.Output))
return nil
}
func marshal(list []v1alpha4.Resource) (buf bytes.Buffer, err error) {
encoder := yaml.NewEncoder(&buf)
defer encoder.Close()
for _, item := range list {
if err = encoder.Encode(item); err != nil {
err = errors.Wrap(err)
return
}
}
return
}
// cacheChart stores a cached copy of Chart in the chart subdirectory of path.
//
// We assume the only method responsible for writing to chartDir is cacheChart
// itself. cacheChart runs concurrently when rendering a platform.
//
// We rely on the atomicity of moving temporary directories into place on the
// same filesystem via os.Rename. If a syscall.EEXIST error occurs during
// renaming, it indicates that the cached chart already exists, which is
// expected when this function is called concurrently.
//
// TODO(jeff): Break the dependency on v1alpha4, make it work across versions as
// a utility function.
func (b *BuildPlan) cacheChart(
ctx context.Context,
log *slog.Logger,
cacheDir string,
chart v1alpha4.Chart,
) error {
// Add repositories
repo := chart.Repository
if repo.URL == "" {
// repo update not needed for oci charts so this is debug instead of warn.
log.DebugContext(ctx, "skipped helm repo add and update: repo url is empty")
} else {
if r, err := util.RunCmd(ctx, "helm", "repo", "add", repo.Name, repo.URL); err != nil {
_, _ = io.Copy(b.Stderr, r.Stderr)
return errors.Format("could not run helm repo add: %w", err)
}
if r, err := util.RunCmd(ctx, "helm", "repo", "update", repo.Name); err != nil {
_, _ = io.Copy(b.Stderr, r.Stderr)
return errors.Format("could not run helm repo update: %w", err)
}
}
cacheTemp, err := os.MkdirTemp(cacheDir, chart.Name)
if err != nil {
return errors.Wrap(err)
}
defer util.Remove(ctx, cacheTemp)
cn := chart.Name
if chart.Repository.Name != "" {
cn = fmt.Sprintf("%s/%s", chart.Repository.Name, chart.Name)
}
helmOut, err := util.RunCmd(ctx, "helm", "pull", "--destination", cacheTemp, "--untar=true", "--version", chart.Version, cn)
if err != nil {
return errors.Wrap(fmt.Errorf("could not run helm pull: %w", err))
}
log.Debug("helm pull", "stdout", helmOut.Stdout, "stderr", helmOut.Stderr)
items, err := os.ReadDir(cacheTemp)
if err != nil {
return errors.Wrap(fmt.Errorf("could not read directory: %w", err))
}
if len(items) != 1 {
return errors.Format("want: exactly one item, have: %+v", items)
}
item := items[0]
src := filepath.Join(cacheTemp, item.Name())
dst := filepath.Join(cacheDir, chart.Name)
if err := os.Rename(src, dst); err != nil {
var linkErr *os.LinkError
if errors.As(err, &linkErr) && errors.Is(linkErr.Err, syscall.EEXIST) {
log.DebugContext(ctx, "cache already exists", "chart", chart.Name, "chart_version", chart.Version, "path", dst)
} else {
return errors.Wrap(fmt.Errorf("could not rename: %w", err))
}
} else {
log.DebugContext(ctx, fmt.Sprintf("renamed %s to %s", src, dst), "src", src, "dst", dst)
}
log.InfoContext(ctx,
fmt.Sprintf("cached %s %s", chart.Name, chart.Version),
"chart", chart.Name,
"chart_version", chart.Version,
"path", dst,
)
return nil
}
// onceWithLock obtains a filesystem lock with mkdir, then executes fn. If the
// lock is already locked, onceWithLock waits for it to be released then returns
// without calling fn.
func onceWithLock(log *slog.Logger, ctx context.Context, path string, fn func() error) error {
if err := os.MkdirAll(filepath.Dir(path), 0777); err != nil {
return errors.Wrap(err)
}
// Obtain a lock with a timeout.
lockDir := path + ".lock"
log = log.With("lock", lockDir)
err := os.Mkdir(lockDir, 0777)
if err == nil {
log.DebugContext(ctx, fmt.Sprintf("acquired %s", lockDir))
defer os.RemoveAll(lockDir)
if err := fn(); err != nil {
return errors.Wrap(err)
}
log.DebugContext(ctx, fmt.Sprintf("released %s", lockDir))
return nil
}
// Wait until the lock is released then return.
if os.IsExist(err) {
log.DebugContext(ctx, fmt.Sprintf("blocked %s", lockDir))
for {
select {
case <-ctx.Done():
return errors.Wrap(ctx.Err())
default:
time.Sleep(100 * time.Millisecond)
if _, err := os.Stat(lockDir); os.IsNotExist(err) {
log.DebugContext(ctx, fmt.Sprintf("unblocked %s", lockDir))
return nil
}
}
}
}
// Unexpected error
return errors.Wrap(err)
}

View File

@@ -20,6 +20,7 @@ func makeBuildRunFunc(cfg *client.Config) command.RunFunc {
ctx := cmd.Root().Context()
logger.FromContext(ctx).DebugContext(ctx, "RunE", "args", args)
build := builder.New(builder.Entrypoints(args), builder.Cluster(cfg.Holos().ClusterName()))
//nolint:staticcheck
results, err := build.Run(ctx, cfg)
if err != nil {
return err

View File

@@ -1,11 +1,18 @@
package render
import (
"bytes"
"context"
"encoding/json"
"flag"
"fmt"
"runtime"
"strings"
h "github.com/holos-run/holos"
"github.com/holos-run/holos/internal/artifact"
"github.com/holos-run/holos/internal/builder"
"github.com/holos-run/holos/internal/builder/v1alpha4"
"github.com/holos-run/holos/internal/cli/command"
"github.com/holos-run/holos/internal/client"
"github.com/holos-run/holos/internal/errors"
@@ -26,8 +33,8 @@ func New(cfg *holos.Config) *cobra.Command {
// New returns the component subcommand for the render command
func NewComponent(cfg *holos.Config) *cobra.Command {
cmd := command.New("component DIRECTORY [DIRECTORY...]")
cmd.Args = cobra.MinimumNArgs(1)
cmd := command.New("component DIRECTORY")
cmd.Args = cobra.ExactArgs(1)
cmd.Short = "render specific components"
cmd.Example = " holos render component --cluster-name=aws2 ./components/monitoring/kube-prometheus-stack"
cmd.Flags().AddGoFlagSet(cfg.WriteFlagSet())
@@ -38,46 +45,102 @@ func NewComponent(cfg *holos.Config) *cobra.Command {
cmd.PersistentFlags().AddGoFlagSet(config.TokenFlagSet())
flagSet := flag.NewFlagSet("", flag.ContinueOnError)
tagMap := make(tags)
flagSet.Var(&tagMap, "tags", "cue tags as comma separated key=value pairs")
var concurrency int
flagSet.IntVar(&concurrency, "concurrency", min(runtime.NumCPU(), 8), "number of concurrent build steps")
cmd.Flags().AddGoFlagSet(flagSet)
cmd.RunE = func(cmd *cobra.Command, args []string) error {
ctx := cmd.Root().Context()
build := builder.New(builder.Entrypoints(args), builder.Cluster(cfg.ClusterName()))
build := builder.New(
builder.Entrypoints(args),
builder.Cluster(cfg.ClusterName()),
builder.Tags(tagMap.Tags()),
)
log := logger.FromContext(ctx)
results, err := build.Run(ctx, config)
log.DebugContext(ctx, "cue: building component instance")
bd, err := build.Unify(ctx, config)
if err != nil {
return errors.Wrap(err)
}
// TODO: Avoid accidental over-writes if two or more holos component
// instances result in the same file path. Write files into a blank
// temporary directory, error if a file exists, then move the directory into
// place.
var result Result
for _, result = range results {
log := logger.FromContext(ctx).With(
"cluster", cfg.ClusterName(),
"name", result.Name(),
)
if result.Continue() {
continue
tm, err := bd.TypeMeta()
if err != nil {
return errors.Wrap(err)
}
if tm.Kind != "BuildPlan" {
return errors.Format("invalid kind: want: BuildPlan have: %s", tm.Kind)
}
log.DebugContext(ctx, "discriminated "+tm.APIVersion+" "+tm.Kind)
jsonBytes, err := bd.Value.MarshalJSON()
if err != nil {
return errors.Format("could not marshal json %s: %w", bd.Dir, err)
}
decoder := json.NewDecoder(bytes.NewReader(jsonBytes))
decoder.DisallowUnknownFields()
art := artifact.New()
switch version := tm.APIVersion; version {
case "v1alpha4":
builder := v1alpha4.BuildPlan{
WriteTo: cfg.WriteTo(),
Concurrency: concurrency,
Stderr: cmd.ErrOrStderr(),
Path: h.InstancePath(args[0]),
}
// DeployFiles from the BuildPlan
if err := result.WriteDeployFiles(ctx, cfg.WriteTo()); err != nil {
if err := decoder.Decode(&builder.BuildPlan); err != nil {
return errors.Format("could not decode build plan %s: %w", bd.Dir, err)
}
return render.Component(ctx, &builder, art)
// Legacy method.
case "v1alpha3", "v1alpha2", "v1alpha1":
//nolint:staticcheck
results, err := build.Run(ctx, config)
if err != nil {
return errors.Wrap(err)
}
// API Objects
if result.SkipWriteAccumulatedOutput() {
log.DebugContext(ctx, "skipped writing k8s objects for "+result.Name())
} else {
path := result.Filename(cfg.WriteTo(), cfg.ClusterName())
if err := result.Save(ctx, path, result.AccumulatedOutput()); err != nil {
// TODO: Avoid accidental over-writes if two or more holos component
// instances result in the same file path. Write files into a blank
// temporary directory, error if a file exists, then move the directory into
// place.
var result Result
for _, result = range results {
log := logger.FromContext(ctx).With(
"cluster", cfg.ClusterName(),
"name", result.Name(),
)
if result.Continue() {
continue
}
// DeployFiles from the BuildPlan
if err := result.WriteDeployFiles(ctx, cfg.WriteTo()); err != nil {
return errors.Wrap(err)
}
// API Objects
if result.SkipWriteAccumulatedOutput() {
log.DebugContext(ctx, "skipped writing k8s objects for "+result.Name())
} else {
path := result.Filename(cfg.WriteTo(), cfg.ClusterName())
if err := result.Save(ctx, path, result.AccumulatedOutput()); err != nil {
return errors.Wrap(err)
}
}
log.InfoContext(ctx, "rendered "+result.Name(), "status", "ok", "action", "rendered")
}
log.InfoContext(ctx, "rendered "+result.Name(), "status", "ok", "action", "rendered")
default:
return errors.Format("component version not supported: %s", version)
}
return nil
}
return cmd
@@ -99,18 +162,87 @@ func NewPlatform(cfg *holos.Config) *cobra.Command {
cmd.RunE = func(cmd *cobra.Command, args []string) error {
ctx := cmd.Root().Context()
build := builder.New(builder.Entrypoints(args))
log := logger.FromContext(ctx)
platform, err := build.Platform(ctx, config)
log.DebugContext(ctx, "cue: building platform instance")
bd, err := build.Unify(ctx, config)
if err != nil {
return errors.Wrap(err)
}
return render.Platform(ctx, concurrency, platform, cmd.ErrOrStderr())
tm, err := bd.TypeMeta()
if err != nil {
return errors.Wrap(err)
}
if tm.Kind != "Platform" {
return errors.Format("invalid kind: want: Platform have: %s", tm.Kind)
}
log.DebugContext(ctx, "discriminated "+tm.APIVersion+" "+tm.Kind)
jsonBytes, err := bd.Value.MarshalJSON()
if err != nil {
return errors.Format("could not marshal json %s: %w", bd.Dir, err)
}
decoder := json.NewDecoder(bytes.NewReader(jsonBytes))
decoder.DisallowUnknownFields()
switch version := tm.APIVersion; version {
case "v1alpha4":
builder := v1alpha4.Platform{
Concurrency: concurrency,
Stderr: cmd.ErrOrStderr(),
}
if err := decoder.Decode(&builder.Platform); err != nil {
return errors.Format("could not decode platform %s: %w", bd.Dir, err)
}
return render.Platform(ctx, &builder)
// Legacy versions prior to the render.Builder interface.
case "v1alpha3", "v1alpha2", "v1alpha1":
platform, err := build.Platform(ctx, config)
if err != nil {
return errors.Wrap(err)
}
//nolint:staticcheck
return render.LegacyPlatform(ctx, concurrency, platform, cmd.ErrOrStderr())
default:
return errors.Format("platform version not supported: %s", version)
}
}
return cmd
}
// tags represents a map of key values for CUE tags for flag parsing.
type tags map[string]string
func (t tags) Tags() []string {
parts := make([]string, 0, len(t))
for k, v := range t {
parts = append(parts, fmt.Sprintf("%s=%s", k, v))
}
return parts
}
func (t tags) String() string {
return strings.Join(t.Tags(), ",")
}
func (t tags) Set(value string) error {
for _, item := range strings.Split(value, ",") {
parts := strings.SplitN(item, "=", 2)
if len(parts) != 2 {
return errors.Format("invalid format, must be tag=value")
}
t[parts[0]] = parts[1]
}
return nil
}
// Deprecated: use render.Artifact instead.
type Result interface {
Continue() bool
Name() string

View File

@@ -13,6 +13,16 @@ import (
// ErrUnsupported is errors.ErrUnsupported
var ErrUnsupported = errors.ErrUnsupported
func NotImplemented() error {
return wrap(New("not implemented"), 2)
}
// Format calls fmt.Format(format, a...) and wraps the error with the caller
// source location.
func Format(format string, a ...any) error {
return wrap(fmt.Errorf(format, a...), 2)
}
// As calls errors.As
func As(err error, target any) bool {
return errors.As(err, target)
@@ -67,11 +77,7 @@ func (e *ErrorAt) Error() string {
return e.Source.Loc() + ": " + e.Err.Error()
}
// Wrap wraps err in a ErrorAt or returns err if err is nil, already a
// ErrorAt, or caller info is not available.
//
// XXX: Refactor to Err(error, ...slog.Attr). Often want to add attributes for the top level logger.
func Wrap(err error) error {
func wrap(err error, skip int) error {
// Nothing to do
if err == nil {
return nil
@@ -84,7 +90,7 @@ func Wrap(err error) error {
}
// Try to wrap err with caller info
if _, file, line, ok := runtime.Caller(1); ok {
if _, file, line, ok := runtime.Caller(skip); ok {
return &ErrorAt{
Err: err,
Source: Source{
@@ -97,6 +103,15 @@ func Wrap(err error) error {
return err
}
// Wrap wraps err in a ErrorAt or returns err if err is nil, already a
// ErrorAt, or caller info is not available.
//
// XXX: Refactor to Err(error, ...slog.Attr). Often want to add attributes for
// the top level logger.
func Wrap(err error) error {
return wrap(err, 2)
}
// Log logs err with Source location if Err is a ErrorAt
func Log(log *slog.Logger, ctx context.Context, level slog.Level, err error, msg string, args ...any) {
var errAt *ErrorAt

View File

@@ -1,4 +1,4 @@
package holos
// Produce a kubectl kustomize build plan.
(#Kustomize & {Name: "{{ .Name }}"}).Output
(#Kustomize & {Name: "{{ .Name }}"}).BuildPlan

View File

@@ -18,4 +18,4 @@ let Objects = {
}
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).Output
(#Kubernetes & Objects).BuildPlan

View File

@@ -8,4 +8,4 @@ let Objects = {
}
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).Output
(#Kubernetes & Objects).BuildPlan

View File

@@ -17,4 +17,4 @@ let Chart = {
}
// Produce a helm chart build plan.
(#Helm & Chart).Output
(#Helm & Chart).BuildPlan

View File

@@ -12,4 +12,4 @@ let Chart = {
}
// Produce a helm chart build plan.
(#Helm & Chart).Output
(#Helm & Chart).BuildPlan

View File

@@ -0,0 +1,11 @@
package holos
// Manage the Component on every Cluster in the Platform
for Fleet in #Fleets {
for Cluster in Fleet.clusters {
#Platform: Components: "\(Cluster.name)/app-projects": {
path: "projects/platform/components/app-projects"
cluster: Cluster.name
}
}
}

View File

@@ -0,0 +1,17 @@
package holos
import ap "argoproj.io/appproject/v1alpha1"
// Registration point for AppProjects
#AppProjects: [Name=string]: ap.#AppProject & {
metadata: name: Name
metadata: namespace: #ArgoCD.Namespace
spec: description: string | *"Holos managed AppProject"
spec: clusterResourceWhitelist: [{group: "*", kind: "*"}]
spec: destinations: [{namespace: "*", server: "*"}]
spec: sourceRepos: ["*"]
}
// Define at least the platform project. Other components can register projects
// the same way from the root of the configuration.
#AppProjects: platform: _

View File

@@ -0,0 +1,9 @@
package holos
let Objects = {
Name: "app-projects"
Resources: AppProject: #AppProjects
}
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan

View File

@@ -0,0 +1,5 @@
{
"name": "app-projects",
"short": "#AppProjects registration point",
"long": "Manage ArgoCD AppProject resources centrally."
}

View File

@@ -0,0 +1,6 @@
package holos
#ArgoConfig: {
Enabled: true
RepoURL: "https://github.com/holos-run/holos-manage-a-project-guide"
}

View File

@@ -0,0 +1,3 @@
package holos
#ArgoConfig: AppProject: #AppProjects.platform.metadata.name

View File

@@ -0,0 +1,5 @@
{
"name": "argocd-config-manage-a-project-guide",
"short": "generate applications for the manage-a-project guide",
"long": "https://github.com/holos-run/holos-manage-a-project-guide"
}

View File

@@ -0,0 +1,15 @@
package holos
// Manage the Component on every Cluster in the Platform
for Fleet in #Fleets {
for Cluster in Fleet.clusters {
#Platform: Components: "\(Cluster.name)/argocd-crds": {
path: "projects/platform/components/argocd/crds"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/argocd": {
path: "projects/platform/components/argocd/argocd"
cluster: Cluster.name
}
}
}

View File

@@ -0,0 +1,13 @@
package holos
// #ArgoCD represents platform wide configuration
#ArgoCD: {
Version: "2.12.3"
Namespace: "argocd"
}
// Register namespaces
#Namespaces: (#ArgoCD.Namespace): _
// Register the HTTPRoute to the backend Service
#HTTPRoutes: argocd: _backendRefs: "argocd-server": namespace: #ArgoCD.Namespace

View File

@@ -0,0 +1,60 @@
package holos
import "strings"
// Produce a helm chart build plan.
(#Helm & Chart).BuildPlan
let Chart = {
Name: "argocd"
Namespace: #ArgoCD.Namespace
Version: "7.5.2"
Repo: name: "argocd"
Repo: url: "https://argoproj.github.io/argo-helm"
Chart: chart: name: "argo-cd"
Chart: chart: release: Name
// Upstream uses a Kubernetes Job to create the argocd-redis Secret. Enable
// hooks to enable the Job.
Chart: enableHooks: true
Resources: [_]: [_]: metadata: namespace: Namespace
// Grant the Gateway namespace the ability to refer to the backend service
// from HTTPRoute resources.
Resources: ReferenceGrant: (#Istio.Gateway.Namespace): #ReferenceGrant
EnableKustomizePostProcessor: true
// Force all resources into the component namespace, some resources in the
// helm chart may not specify the namespace so they may get mis-applied
// depending on the kubectl (client-go) context.
KustomizeFiles: "kustomization.yaml": namespace: Namespace
Values: #Values & {
kubeVersionOverride: "1.29.0"
// handled in the argo-crds component
crds: install: false
// Configure the same fqdn the HTTPRoute is configured with.
global: domain: #HTTPRoutes.argocd.spec.hostnames[0]
dex: enabled: false
// the platform handles mutual tls to the backend
configs: params: "server.insecure": true
configs: cm: {
"admin.enabled": false
"oidc.config": "{}"
"users.anonymous.enabled": "true"
}
// Refer to https://argo-cd.readthedocs.io/en/stable/operator-manual/rbac/
let Policy = [
"g, argocd-view, role:readonly",
"g, prod-cluster-view, role:readonly",
"g, prod-cluster-edit, role:readonly",
"g, prod-cluster-admin, role:admin",
]
configs: rbac: "policy.csv": strings.Join(Policy, "\n")
configs: rbac: "policy.default": "role:admin"
}
}

View File

@@ -0,0 +1,37 @@
package holos
import (
"encoding/yaml"
ks "sigs.k8s.io/kustomize/api/types"
)
(#Kubernetes & {Name: "argocd-crds"}).BuildPlan
// Holos stages BuildPlan resources as an intermediate step of the rendering
// pipeline. The purpose is to provide the resources to kustomize for
// post-processing.
let BuildPlanResources = "build-plan-resources.yaml"
let Kustomization = ks.#Kustomization & {
apiVersion: "kustomize.config.k8s.io/v1beta1"
kind: "Kustomization"
resources: [
// Kustomize the intermediate build plan resources.
BuildPlanResources,
// Mix-in external resources.
"https://raw.githubusercontent.com/argoproj/argo-cd/v\(#ArgoCD.Version)/manifests/crds/application-crd.yaml",
"https://raw.githubusercontent.com/argoproj/argo-cd/v\(#ArgoCD.Version)/manifests/crds/applicationset-crd.yaml",
"https://raw.githubusercontent.com/argoproj/argo-cd/v\(#ArgoCD.Version)/manifests/crds/appproject-crd.yaml",
// This method also works, but takes about 5 seconds longer each build.
// "https://github.com/argoproj/argo-cd//manifests/crds/?ref=v\(#ArgoCD.Version)",
]
}
// Generate a kustomization.yaml directly from CUE so we can provide the correct
// version.
spec: components: kubernetesObjectsList: [{
// intermediate build plan resources to kustomize. Necessary to activate the
// kustomization post-rendering step in holos.
kustomize: resourcesFile: BuildPlanResources
kustomize: kustomizeFiles: "kustomization.yaml": yaml.Marshal(Kustomization)
}]

View File

@@ -0,0 +1,5 @@
{
"name": "argocd",
"short": "declaritive gitops for kubernetes",
"long": "Argo CD is a declarative, GitOps continuous delivery tool for Kubernetes."
}

View File

@@ -0,0 +1,51 @@
package holos
// Manage on workload clusters only
for Cluster in #Fleets.workload.clusters {
// Owned by the security team
#Platform: Components: "\(Cluster.name)/bank-secrets": {
path: "projects/bank-of-holos/security/components/bank-secrets"
cluster: Cluster.name
}
// Owned by the frontend team
#Platform: Components: "\(Cluster.name)/bank-frontend": {
path: "projects/bank-of-holos/frontend/components/bank-frontend"
cluster: Cluster.name
}
// Owned by the backend team
#Platform: Components: "\(Cluster.name)/bank-backend-config": {
path: "projects/bank-of-holos/backend/components/bank-backend-config"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/bank-accounts-db": {
path: "projects/bank-of-holos/backend/components/bank-accounts-db"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/bank-userservice": {
path: "projects/bank-of-holos/backend/components/bank-userservice"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/bank-ledger-db": {
path: "projects/bank-of-holos/backend/components/bank-ledger-db"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/bank-ledger-writer": {
path: "projects/bank-of-holos/backend/components/bank-ledger-writer"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/bank-balance-reader": {
path: "projects/bank-of-holos/backend/components/bank-balance-reader"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/bank-transaction-history": {
path: "projects/bank-of-holos/backend/components/bank-transaction-history"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/bank-contacts": {
path: "projects/bank-of-holos/backend/components/bank-contacts"
cluster: Cluster.name
}
}

View File

@@ -0,0 +1,26 @@
package holos
// Platform wide definitions
#BankOfHolos: {
Name: "bank-of-holos"
Frontend: Namespace: "bank-frontend"
Backend: Namespace: "bank-backend"
Security: Namespace: "bank-security"
// Resources to manage in each of the namespaces.
Resources: #Resources
}
// Register namespaces
#Namespaces: (#BankOfHolos.Frontend.Namespace): _
#Namespaces: (#BankOfHolos.Backend.Namespace): _
#Namespaces: (#BankOfHolos.Security.Namespace): _
// Register projects
#AppProjects: "bank-frontend": _
#AppProjects: "bank-backend": _
#AppProjects: "bank-security": _
// Register HTTPRoutes.
// bank.example.com routes to Service frontend in the bank-frontend namespace.
#HTTPRoutes: bank: _backendRefs: frontend: namespace: #BankOfHolos.Frontend.Namespace

View File

@@ -0,0 +1,3 @@
package holos
#ArgoConfig: AppProject: #AppProjects["bank-backend"].metadata.name

View File

@@ -0,0 +1,122 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
team: "accounts"
tier: "db"
}
let Objects = {
Name: "bank-accounts-db"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests/accounts-db.yaml
Resources: {
ConfigMap: "accounts-db-config": {
apiVersion: "v1"
data: {
ACCOUNTS_DB_URI: "postgresql://accounts-admin:accounts-pwd@accounts-db:5432/accounts-db"
POSTGRES_DB: "accounts-db"
POSTGRES_PASSWORD: "accounts-pwd"
POSTGRES_USER: "accounts-admin"
}
kind: "ConfigMap"
metadata: {
labels: {
app: "accounts-db"
CommonLabels
}
name: "accounts-db-config"
}
}
Service: "accounts-db": {
apiVersion: "v1"
kind: "Service"
metadata: {
name: "accounts-db"
labels: CommonLabels
}
spec: {
ports: [{
name: "tcp"
port: 5432
protocol: "TCP"
targetPort: 5432
}]
selector: {
app: "accounts-db"
CommonLabels
}
type: "ClusterIP"
}
}
StatefulSet: "accounts-db": {
apiVersion: "apps/v1"
kind: "StatefulSet"
metadata: {
name: "accounts-db"
labels: CommonLabels
}
spec: {
replicas: 1
selector: matchLabels: {
app: "accounts-db"
CommonLabels
}
serviceName: "accounts-db"
template: {
metadata: labels: {
app: "accounts-db"
CommonLabels
}
spec: {
containers: [{
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "accounts-db-config"
}, {
configMapRef: name: "demo-data-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/accounts-db:v0.6.5@sha256:abb955756a82b115e0fd9c5fa1527ae1a744b398b357fd6d7a26348feccad181"
name: "accounts-db"
ports: [{containerPort: 5432}]
resources: {
limits: {
cpu: "250m"
memory: "512Mi"
}
requests: {
cpu: "100m"
memory: "128Mi"
}
}
volumeMounts: [{
mountPath: "/var/lib/postgresql/data"
name: "postgresdb"
subPath: "postgres"
}]
}]
serviceAccount: BankName
serviceAccountName: BankName
volumes: [{
emptyDir: {}
name: "postgresdb"
}]
}
}
}
}
}
}

View File

@@ -0,0 +1,33 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
tier: "backend"
}
let Objects = {
Name: "bank-backend-config"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
Resources: [_]: [_]: metadata: labels: CommonLabels
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests/userservice.yaml
Resources: {
// Allow HTTPRoutes in the ingress gateway namespace to reference Services
// in this namespace.
ReferenceGrant: grant: #ReferenceGrant & {
metadata: namespace: Namespace
}
// Include shared resources
#BankOfHolos.Resources
}
}

View File

@@ -0,0 +1,179 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
team: "ledger"
tier: "backend"
}
let Objects = {
Name: "bank-balance-reader"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
Resources: [_]: [_]: metadata: labels: CommonLabels
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests
Resources: {
Service: balancereader: {
apiVersion: "v1"
kind: "Service"
metadata: {
name: "balancereader"
labels: CommonLabels
}
spec: {
ports: [{
name: "http"
port: 8080
targetPort: 8080
}]
selector: {
app: "balancereader"
CommonLabels
}
type: "ClusterIP"
}
}
Deployment: balancereader: {
apiVersion: "apps/v1"
kind: "Deployment"
metadata: {
name: "balancereader"
labels: CommonLabels
}
spec: {
selector: matchLabels: {
app: "balancereader"
CommonLabels
}
template: {
metadata: labels: {
app: "balancereader"
CommonLabels
}
spec: {
containers: [{
env: [{
name: "VERSION"
value: "v0.6.5"
}, {
name: "PORT"
value: "8080"
}, {
name: "ENABLE_TRACING"
value: "false"
}, {
name: "ENABLE_METRICS"
value: "false"
}, {
name: "POLL_MS"
value: "100"
}, {
name: "CACHE_SIZE"
value: "1000000"
}, {
name: "JVM_OPTS"
value: "-XX:+UnlockExperimentalVMOptions -XX:+UseCGroupMemoryLimitForHeap -Xms256m -Xmx512m"
}, {
name: "LOG_LEVEL"
value: "info"
}, {
name: "NAMESPACE"
valueFrom: fieldRef: fieldPath: "metadata.namespace"
}]
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "ledger-db-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/balancereader:v0.6.5@sha256:de01f16554ae2d0b49ac85116e6307da8c0f8a35f50a0cf25e1e4a4fe18dca83"
livenessProbe: {
httpGet: {
path: "/healthy"
port: 8080
}
initialDelaySeconds: 120
periodSeconds: 5
timeoutSeconds: 10
}
name: "balancereader"
readinessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 60
periodSeconds: 5
timeoutSeconds: 10
}
resources: {
limits: {
cpu: "500m"
"ephemeral-storage": "0.5Gi"
memory: "512Mi"
}
requests: {
cpu: "100m"
"ephemeral-storage": "0.5Gi"
memory: "256Mi"
}
}
securityContext: {
allowPrivilegeEscalation: false
capabilities: drop: ["all"]
privileged: false
readOnlyRootFilesystem: true
}
startupProbe: {
failureThreshold: 30
httpGet: {
path: "/healthy"
port: 8080
}
periodSeconds: 10
}
volumeMounts: [{
mountPath: "/tmp"
name: "tmp"
}, {
mountPath: "/tmp/.ssh"
name: "publickey"
readOnly: true
}]
}]
securityContext: {
fsGroup: 1000
runAsGroup: 1000
runAsNonRoot: true
runAsUser: 1000
}
serviceAccountName: BankName
terminationGracePeriodSeconds: 5
volumes: [{
emptyDir: {}
name: "tmp"
}, {
name: "publickey"
secret: {
items: [{
key: "jwtRS256.key.pub"
path: "publickey"
}]
secretName: "jwt-key"
}
}]
}
}
}
}
}
}

View File

@@ -0,0 +1,143 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
team: "accounts"
tier: "backend"
}
let Objects = {
Name: "bank-contacts"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
Resources: [_]: [_]: metadata: labels: CommonLabels
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests
Resources: {
Service: contacts: {
apiVersion: "v1"
kind: "Service"
metadata: name: "contacts"
spec: {
ports: [{
name: "http"
port: 8080
targetPort: 8080
}]
selector: {
app: "contacts"
CommonLabels
}
type: "ClusterIP"
}
}
Deployment: contacts: {
apiVersion: "apps/v1"
kind: "Deployment"
metadata: name: "contacts"
spec: {
selector: matchLabels: {
app: "contacts"
CommonLabels
}
template: {
metadata: {
labels: {
app: "contacts"
CommonLabels
}
}
spec: {
containers: [{
env: [{
name: "VERSION"
value: "v0.6.5"
}, {
name: "PORT"
value: "8080"
}, {
name: "ENABLE_TRACING"
value: "false"
}, {
name: "LOG_LEVEL"
value: "info"
}]
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "accounts-db-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/contacts:v0.6.5@sha256:e451dcac7d34a7bde979c7f02d4c7ebd83a77aff373e1131ce3a2bba2f7fdc1a"
name: "contacts"
readinessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 10
periodSeconds: 5
timeoutSeconds: 10
}
resources: {
limits: {
cpu: "250m"
"ephemeral-storage": "0.25Gi"
memory: "128Mi"
}
requests: {
cpu: "100m"
"ephemeral-storage": "0.25Gi"
memory: "64Mi"
}
}
securityContext: {
allowPrivilegeEscalation: false
capabilities: drop: ["all"]
privileged: false
readOnlyRootFilesystem: true
}
volumeMounts: [{
mountPath: "/tmp"
name: "tmp"
}, {
mountPath: "/tmp/.ssh"
name: "publickey"
readOnly: true
}]
}]
securityContext: {
fsGroup: 1000
runAsGroup: 1000
runAsNonRoot: true
runAsUser: 1000
}
serviceAccountName: BankName
terminationGracePeriodSeconds: 5
volumes: [{
emptyDir: {}
name: "tmp"
}, {
name: "publickey"
secret: {
items: [{
key: "jwtRS256.key.pub"
path: "publickey"
}]
secretName: "jwt-key"
}
}]
}
}
}
}
}
}

View File

@@ -0,0 +1,121 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
team: "ledger"
tier: "db"
}
let Objects = {
Name: "bank-ledger-db"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests
Resources: {
ConfigMap: "ledger-db-config": {
apiVersion: "v1"
metadata: {
name: "ledger-db-config"
labels: {
app: "ledger-db"
CommonLabels
}
}
data: {
POSTGRES_DB: "postgresdb"
POSTGRES_PASSWORD: "password"
POSTGRES_USER: "admin"
SPRING_DATASOURCE_PASSWORD: "password"
SPRING_DATASOURCE_URL: "jdbc:postgresql://ledger-db:5432/postgresdb"
SPRING_DATASOURCE_USERNAME: "admin"
}
}
Service: "ledger-db": {
apiVersion: "v1"
kind: "Service"
metadata: {
name: "ledger-db"
labels: CommonLabels
}
spec: {
ports: [{
name: "tcp"
port: 5432
targetPort: 5432
}]
selector: {
app: "ledger-db"
CommonLabels
}
type: "ClusterIP"
}
}
StatefulSet: "ledger-db": {
apiVersion: "apps/v1"
kind: "StatefulSet"
metadata: {
name: "ledger-db"
labels: CommonLabels
}
spec: {
replicas: 1
selector: matchLabels: {
app: "ledger-db"
CommonLabels
}
serviceName: "ledger-db"
template: {
metadata: labels: {
app: "ledger-db"
CommonLabels
}
spec: {
containers: [{
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "ledger-db-config"
}, {
configMapRef: name: "demo-data-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/ledger-db:v0.6.5@sha256:cc4fd25f301ab6d46b1312244d6931babc4c6cb66c5cb6d31d4a1adfa318a321"
name: "postgres"
ports: [{containerPort: 5432}]
resources: {
limits: {
cpu: "250m"
memory: "1Gi"
}
requests: {
cpu: "100m"
memory: "512Mi"
}
}
volumeMounts: [{
mountPath: "/var/lib/postgresql/data"
name: "postgresdb"
subPath: "postgres"
}]
}]
serviceAccountName: BankName
volumes: [{
emptyDir: {}
name: "postgresdb"
}]
}
}
}
}
}
}

View File

@@ -0,0 +1,169 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
team: "ledger"
tier: "backend"
}
let Objects = {
Name: "bank-ledger-writer"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
Resources: [_]: [_]: metadata: labels: CommonLabels
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests
Resources: {
Service: ledgerwriter: {
apiVersion: "v1"
kind: "Service"
metadata: {
name: "ledgerwriter"
labels: CommonLabels
}
spec: {
ports: [{
name: "http"
port: 8080
targetPort: 8080
}]
selector: {
app: "ledgerwriter"
CommonLabels
}
type: "ClusterIP"
}
}
Deployment: ledgerwriter: {
apiVersion: "apps/v1"
kind: "Deployment"
metadata: {
name: "ledgerwriter"
labels: CommonLabels
}
spec: {
selector: matchLabels: {
app: "ledgerwriter"
CommonLabels
}
template: {
metadata: {
labels: {
app: "ledgerwriter"
CommonLabels
}
}
spec: {
containers: [{
env: [{
name: "VERSION"
value: "v0.6.5"
}, {
name: "PORT"
value: "8080"
}, {
name: "ENABLE_TRACING"
value: "false"
}, {
name: "ENABLE_METRICS"
value: "false"
}, {
name: "JVM_OPTS"
value: "-XX:+UnlockExperimentalVMOptions -XX:+UseCGroupMemoryLimitForHeap -Xms256m -Xmx512m"
}, {
name: "LOG_LEVEL"
value: "info"
}, {
name: "NAMESPACE"
valueFrom: fieldRef: fieldPath: "metadata.namespace"
}]
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "service-api-config"
}, {
configMapRef: name: "ledger-db-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/ledgerwriter:v0.6.5@sha256:5b66d6888b87993c8ebe260fe33005c4e4bc2bdae4b5682874e1a078d37ff3b2"
name: "ledgerwriter"
readinessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 60
periodSeconds: 5
timeoutSeconds: 10
}
resources: {
limits: {
cpu: "500m"
"ephemeral-storage": "0.5Gi"
memory: "512Mi"
}
requests: {
cpu: "100m"
"ephemeral-storage": "0.5Gi"
memory: "256Mi"
}
}
securityContext: {
allowPrivilegeEscalation: false
capabilities: drop: ["all"]
privileged: false
readOnlyRootFilesystem: true
}
startupProbe: {
failureThreshold: 30
httpGet: {
path: "/ready"
port: 8080
}
periodSeconds: 10
}
volumeMounts: [{
mountPath: "/tmp"
name: "tmp"
}, {
mountPath: "/tmp/.ssh"
name: "publickey"
readOnly: true
}]
}]
securityContext: {
fsGroup: 1000
runAsGroup: 1000
runAsNonRoot: true
runAsUser: 1000
}
serviceAccountName: BankName
terminationGracePeriodSeconds: 5
volumes: [{
emptyDir: {}
name: "tmp"
}, {
name: "publickey"
secret: {
items: [{
key: "jwtRS256.key.pub"
path: "publickey"
}]
secretName: "jwt-key"
}
}]
}
}
}
}
}
}

View File

@@ -0,0 +1,187 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
team: "ledger"
tier: "backend"
}
let Objects = {
Name: "bank-transaction-history"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
Resources: [_]: [_]: metadata: labels: CommonLabels
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests
Resources: {
Service: transactionhistory: {
apiVersion: "v1"
kind: "Service"
metadata: {
labels: CommonLabels
name: "transactionhistory"
}
spec: {
ports: [{
name: "http"
port: 8080
targetPort: 8080
}]
selector: {
app: "transactionhistory"
CommonLabels
}
type: "ClusterIP"
}
}
Deployment: transactionhistory: {
apiVersion: "apps/v1"
kind: "Deployment"
metadata: {
name: "transactionhistory"
labels: CommonLabels
}
spec: {
selector: matchLabels: {
app: "transactionhistory"
CommonLabels
}
template: {
metadata: {
labels: {
app: "transactionhistory"
CommonLabels
}
}
spec: {
containers: [{
env: [{
name: "VERSION"
value: "v0.6.5"
}, {
name: "PORT"
value: "8080"
}, {
name: "ENABLE_TRACING"
value: "false"
}, {
name: "ENABLE_METRICS"
value: "false"
}, {
name: "POLL_MS"
value: "100"
}, {
name: "CACHE_SIZE"
value: "1000"
}, {
name: "CACHE_MINUTES"
value: "60"
}, {
name: "HISTORY_LIMIT"
value: "100"
}, {
name: "JVM_OPTS"
value: "-XX:+UnlockExperimentalVMOptions -XX:+UseCGroupMemoryLimitForHeap -Xms256m -Xmx512m"
}, {
name: "LOG_LEVEL"
value: "info"
}, {
name: "NAMESPACE"
valueFrom: fieldRef: fieldPath: "metadata.namespace"
}]
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "ledger-db-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/transactionhistory:v0.6.5@sha256:54a2b0866df44a50832e71b130f3e069fe8bbce71309fb6cf390b19f64d92c09"
livenessProbe: {
httpGet: {
path: "/healthy"
port: 8080
}
initialDelaySeconds: 120
periodSeconds: 5
timeoutSeconds: 10
}
name: "transactionhistory"
readinessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 60
periodSeconds: 5
timeoutSeconds: 10
}
resources: {
limits: {
cpu: "500m"
"ephemeral-storage": "0.5Gi"
memory: "512Mi"
}
requests: {
cpu: "100m"
"ephemeral-storage": "0.5Gi"
memory: "256Mi"
}
}
securityContext: {
allowPrivilegeEscalation: false
capabilities: drop: ["all"]
privileged: false
readOnlyRootFilesystem: true
}
startupProbe: {
failureThreshold: 30
httpGet: {
path: "/healthy"
port: 8080
}
periodSeconds: 10
}
volumeMounts: [{
mountPath: "/tmp"
name: "tmp"
}, {
mountPath: "/tmp/.ssh"
name: "publickey"
readOnly: true
}]
}]
securityContext: {
fsGroup: 1000
runAsGroup: 1000
runAsNonRoot: true
runAsUser: 1000
}
serviceAccountName: BankName
terminationGracePeriodSeconds: 5
volumes: [{
emptyDir: {}
name: "tmp"
}, {
name: "publickey"
secret: {
items: [{
key: "jwtRS256.key.pub"
path: "publickey"
}]
secretName: "jwt-key"
}
}]
}
}
}
}
}
}

View File

@@ -0,0 +1,156 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let BankName = #BankOfHolos.Name
let CommonLabels = {
application: BankName
environment: "development"
team: "accounts"
tier: "backend"
}
let Objects = {
Name: "bank-userservice"
Namespace: #BankOfHolos.Backend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests/userservice.yaml
Resources: {
Service: userservice: {
metadata: name: "userservice"
metadata: labels: CommonLabels
spec: {
selector: {
app: "userservice"
CommonLabels
}
_ports: http: {
name: "http"
port: 8080
targetPort: 8080
protocol: "TCP"
}
ports: [for x in _ports {x}]
}
}
Deployment: userservice: {
metadata: name: "userservice"
metadata: labels: CommonLabels
spec: {
selector: matchLabels: {
app: "userservice"
CommonLabels
}
template: {
metadata: labels: {
app: "userservice"
CommonLabels
}
spec: {
serviceAccountName: BankName
terminationGracePeriodSeconds: 5
containers: [{
env: [{
name: "VERSION"
value: "v0.6.5"
}, {
name: "PORT"
value: "8080"
}, {
name: "ENABLE_TRACING"
value: "false"
}, {
name: "LOG_LEVEL"
value: "info"
}, {
name: "TOKEN_EXPIRY_SECONDS"
value: "3600"
}, {
name: "PRIV_KEY_PATH"
value: "/tmp/.ssh/privatekey"
}]
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "accounts-db-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/userservice:v0.6.5@sha256:f91e0e5bd6cdb16f6b867b2e3e874b23dd01f11592de006776f1dfb136702941"
name: "userservice"
ports: [{
containerPort: 8080
name: "http-server"
}]
readinessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 10
periodSeconds: 5
timeoutSeconds: 10
}
resources: {
limits: {
cpu: "500m"
"ephemeral-storage": "0.25Gi"
memory: "256Mi"
}
requests: {
cpu: "260m"
"ephemeral-storage": "0.25Gi"
memory: "128Mi"
}
}
securityContext: {
allowPrivilegeEscalation: false
capabilities: drop: ["all"]
privileged: false
readOnlyRootFilesystem: true
}
volumeMounts: [{
mountPath: "/tmp"
name: "tmp"
}, {
mountPath: "/tmp/.ssh"
name: "keys"
readOnly: true
}]
}]
volumes: [{
emptyDir: {}
name: "tmp"
}, {
name: "keys"
secret: {
secretName: "jwt-key"
items: [
{
key: "jwtRS256.key"
path: "privatekey"
},
{
key: "jwtRS256.key.pub"
path: "publickey"
},
]
}
}]
securityContext: {
seccompProfile: type: "RuntimeDefault"
fsGroup: 1000
runAsGroup: 1000
runAsNonRoot: true
runAsUser: 1000
}
}
}
}
}
}
}

View File

@@ -0,0 +1,88 @@
package holos
import (
core "k8s.io/api/core/v1"
es "external-secrets.io/externalsecret/v1beta1"
ss "external-secrets.io/secretstore/v1beta1"
)
let BankName = #BankOfHolos.Name
#BankOfHolos: {
// Resources to make available in each of the project namespaces.
Resources: {
ServiceAccount: (BankName): core.#ServiceAccount & {
apiVersion: "v1"
kind: "ServiceAccount"
metadata: name: BankName
}
// SecretStore to fetch secrets owned by the security team
SecretStore: (BankName): ss.#SecretStore & {
metadata: name: #BankOfHolos.Security.Namespace
spec: provider: {
kubernetes: {
remoteNamespace: #BankOfHolos.Security.Namespace
auth: serviceAccount: name: ServiceAccount[BankName].metadata.name
server: {
url: "https://kubernetes.default.svc"
caProvider: {
type: "ConfigMap"
name: "kube-root-ca.crt"
key: "ca.crt"
}
}
}
}
}
// We do not check the private key into version control.
// https://github.com/GoogleCloudPlatform/bank-of-anthos/tree/v0.6.5/extras/jwt
ExternalSecret: "jwt-key": es.#ExternalSecret & {
metadata: name: "jwt-key"
spec: {
target: name: metadata.name
dataFrom: [{extract: {key: metadata.name}}]
refreshInterval: "5s"
secretStoreRef: kind: "SecretStore"
secretStoreRef: name: SecretStore[BankName].metadata.name
}
}
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests/config.yaml
ConfigMap: "environment-config": core.#ConfigMap & {
apiVersion: "v1"
kind: "ConfigMap"
metadata: name: "environment-config"
data: {
LOCAL_ROUTING_NUM: "883745000"
PUB_KEY_PATH: "/tmp/.ssh/publickey"
}
}
ConfigMap: "service-api-config": core.#ConfigMap & {
apiVersion: "v1"
kind: "ConfigMap"
metadata: name: "service-api-config"
data: {
TRANSACTIONS_API_ADDR: "ledgerwriter.\(#BankOfHolos.Backend.Namespace).svc:8080"
BALANCES_API_ADDR: "balancereader.\(#BankOfHolos.Backend.Namespace).svc:8080"
HISTORY_API_ADDR: "transactionhistory.\(#BankOfHolos.Backend.Namespace).svc:8080"
CONTACTS_API_ADDR: "contacts.\(#BankOfHolos.Backend.Namespace).svc:8080"
USERSERVICE_API_ADDR: "userservice.\(#BankOfHolos.Backend.Namespace).svc:8080"
}
}
ConfigMap: "demo-data-config": core.#ConfigMap & {
apiVersion: "v1"
kind: "ConfigMap"
metadata: name: "demo-data-config"
data: {
USE_DEMO_DATA: "True"
DEMO_LOGIN_USERNAME: "testuser"
// All demo user accounts are hardcoded to use the login password 'bankofanthos'
DEMO_LOGIN_PASSWORD: "bankofanthos"
}
}
}
}

View File

@@ -0,0 +1,201 @@
package holos
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
let Objects = {
Name: "bank-frontend"
Namespace: #BankOfHolos.Frontend.Namespace
// Ensure resources go in the correct namespace
Resources: [_]: [_]: metadata: namespace: Namespace
// https://github.com/GoogleCloudPlatform/bank-of-anthos/blob/release/v0.6.5/kubernetes-manifests/frontend.yaml
Resources: {
Service: frontend: {
metadata: name: "frontend"
metadata: labels: {
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
spec: {
selector: {
app: "frontend"
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
_ports: http: {
name: "http"
port: 80
targetPort: 8080
protocol: "TCP"
}
ports: [for x in _ports {x}]
}
}
Deployment: frontend: {
metadata: name: "frontend"
metadata: labels: {
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
spec: {
selector: matchLabels: {
app: "frontend"
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
template: {
metadata: labels: {
app: "frontend"
application: "bank-of-holos"
environment: "development"
team: "frontend"
tier: "web"
}
spec: {
securityContext: {
seccompProfile: type: "RuntimeDefault"
fsGroup: 1000
runAsGroup: 1000
runAsNonRoot: true
runAsUser: 1000
}
serviceAccountName: "bank-of-holos"
terminationGracePeriodSeconds: 5
containers: [{
env: [{
name: "BANK_NAME"
value: "Bank of Holos"
}, {
name: "ENV_PLATFORM"
value: "local"
}, {
name: "VERSION"
value: "v0.6.5"
}, {
name: "PORT"
value: "8080"
}, {
name: "ENABLE_TRACING"
value: "false"
}, {
name: "SCHEME"
value: "https"
}, {
name: "LOG_LEVEL"
value: "info"
}, {
name: "DEFAULT_USERNAME"
valueFrom: configMapKeyRef: {
key: "DEMO_LOGIN_USERNAME"
name: "demo-data-config"
}
}, {
name: "DEFAULT_PASSWORD"
valueFrom: configMapKeyRef: {
key: "DEMO_LOGIN_PASSWORD"
name: "demo-data-config"
}
}, {
name: "REGISTERED_OAUTH_CLIENT_ID"
valueFrom: configMapKeyRef: {
key: "DEMO_OAUTH_CLIENT_ID"
name: "oauth-config"
optional: true
}
}, {
name: "ALLOWED_OAUTH_REDIRECT_URI"
valueFrom: configMapKeyRef: {
key: "DEMO_OAUTH_REDIRECT_URI"
name: "oauth-config"
optional: true
}
}]
envFrom: [{
configMapRef: name: "environment-config"
}, {
configMapRef: name: "service-api-config"
}]
image: "us-central1-docker.pkg.dev/bank-of-anthos-ci/bank-of-anthos/frontend:v0.6.5@sha256:d72050f70d12383e4434ad04d189b681dc625f696087ddf0b5df641645c9dafa"
livenessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 60
periodSeconds: 15
timeoutSeconds: 30
}
name: "front"
readinessProbe: {
httpGet: {
path: "/ready"
port: 8080
}
initialDelaySeconds: 10
periodSeconds: 5
timeoutSeconds: 10
}
resources: {
limits: {
cpu: "250m"
memory: "128Mi"
}
requests: {
cpu: "100m"
memory: "64Mi"
}
}
securityContext: {
allowPrivilegeEscalation: false
capabilities: drop: ["all"]
privileged: false
readOnlyRootFilesystem: true
}
volumeMounts: [{
mountPath: "/tmp"
name: "tmp"
}, {
mountPath: "/tmp/.ssh"
name: "publickey"
readOnly: true
}]
}]
volumes: [
{
emptyDir: {}
name: "tmp"
},
{
name: "publickey"
secret: {
items: [{key: "jwtRS256.key.pub", path: "publickey"}]
secretName: "jwt-key"
}
},
]
}
}
}
}
// Allow HTTPRoutes in the ingress gateway namespace to reference Services
// in this namespace.
ReferenceGrant: grant: #ReferenceGrant & {
metadata: namespace: Namespace
}
// Include shared resources
#BankOfHolos.Resources
}
}

View File

@@ -0,0 +1,3 @@
package holos
#ArgoConfig: AppProject: #AppProjects["bank-frontend"].metadata.name

View File

@@ -0,0 +1,165 @@
package holos
import (
corev1 "k8s.io/api/core/v1"
rbacv1 "k8s.io/api/rbac/v1"
batchv1 "k8s.io/api/batch/v1"
)
// Produce a kubernetes objects build plan.
(#Kubernetes & Objects).BuildPlan
// This may be useful to copy and generate other secrets.
let SecretName = "jwt-key"
// Roles for reading and writing secrets
let Reader = "\(SecretName)-reader"
let Writer = "\(SecretName)-writer"
// AllowedName represents the service account allowed to read the generated
// secret.
let AllowedName = #BankOfHolos.Name
let Objects = {
Name: "bank-secrets"
Namespace: #BankOfHolos.Security.Namespace
Resources: [_]: [_]: metadata: namespace: Namespace
Resources: [_]: [ID=string]: metadata: name: string | *ID
Resources: {
// Kubernetes ServiceAccount used by the secret generator job.
ServiceAccount: (Writer): corev1.#ServiceAccount
// Role to allow the ServiceAccount to update secrets.
Role: (Writer): rbacv1.#Role & {
rules: [{
apiGroups: [""]
resources: ["secrets"]
verbs: ["create", "update", "patch"]
}]
}
// Bind the role to the service account.
RoleBinding: (Writer): rbacv1.#RoleBinding & {
roleRef: {
apiGroup: "rbac.authorization.k8s.io"
kind: "Role"
name: Role[Writer].metadata.name
}
subjects: [{
kind: "ServiceAccount"
name: ServiceAccount[Writer].metadata.name
namespace: Namespace
}]
}
let JobSpec = {
serviceAccountName: Writer
restartPolicy: "OnFailure"
securityContext: {
seccompProfile: type: "RuntimeDefault"
runAsNonRoot: true
runAsUser: 8192 // app
}
containers: [
{
name: "toolkit"
image: "quay.io/holos-run/toolkit:2024-09-16"
securityContext: {
capabilities: drop: ["ALL"]
allowPrivilegeEscalation: false
}
command: ["/bin/bash"]
args: ["/config/entrypoint"]
env: [{
name: "HOME"
value: "/tmp"
}]
volumeMounts: [{
name: "config"
mountPath: "/config"
readOnly: true
}]
},
]
volumes: [{
name: "config"
configMap: name: Writer
}]
}
Job: (Writer): batchv1.#Job & {
spec: template: spec: JobSpec
}
ConfigMap: (Writer): corev1.#ConfigMap & {
data: entrypoint: ENTRYPOINT
}
// Allow the SecretStore in the frontend and backend namespaces to read the
// secret.
Role: (Reader): rbacv1.#Role & {
rules: [{
apiGroups: [""]
resources: ["secrets"]
resourceNames: [SecretName]
verbs: ["get"]
}]
}
// Grant access to the bank-of-holos service account in the frontend and
// backend namespaces.
RoleBinding: (Reader): rbacv1.#RoleBinding & {
roleRef: {
apiGroup: "rbac.authorization.k8s.io"
kind: "Role"
name: Role[Reader].metadata.name
}
subjects: [{
kind: "ServiceAccount"
name: AllowedName
namespace: #BankOfHolos.Frontend.Namespace
}, {
kind: "ServiceAccount"
name: AllowedName
namespace: #BankOfHolos.Backend.Namespace
},
]
}
}
}
let ENTRYPOINT = """
#! /bin/bash
#
tmpdir="$(mktemp -d)"
finish() {
status=$?
rm -rf "${tmpdir}"
return $status
}
trap finish EXIT
set -euo pipefail
cd "$tmpdir"
mkdir secret
cd secret
echo "generating private key" >&2
ssh-keygen -t rsa -b 4096 -m PEM -f jwtRS256.key -q -N "" -C \(AllowedName)
echo "generating public key" >&2
ssh-keygen -e -m PKCS8 -f jwtRS256.key > jwtRS256.key.pub
cd ..
echo "copying secret into kubernetes manifest secret.yaml" >&2
kubectl create secret generic \(SecretName) --from-file=secret --dry-run=client -o yaml > secret.yaml
echo "applying secret.yaml" >&2
kubectl apply --server-side=true -f secret.yaml
echo "cleaning up" >&2
rm -rf secret secret.yaml
echo "ok done" >&2
"""

View File

@@ -0,0 +1,3 @@
package holos
#ArgoConfig: AppProject: #AppProjects["bank-security"].metadata.name

View File

@@ -0,0 +1,5 @@
{
"name": "bank-of-holos",
"short": "demo bank composed of two projects",
"long": "Bank of Holos is a sample HTTP-based web app that simulates a bank's payment processing network, allowing users to create artificial bank accounts and complete transactions."
}

View File

@@ -0,0 +1,20 @@
package holos
// Platform wide configuration
#CertManager: {
Version: "{{ .Version }}"
Namespace: "{{ .Namespace }}"
}
// Register the namespace
#Namespaces: (#CertManager.Namespace): _
// Manage the component on every cluster in the platform
for Fleet in #Fleets {
for Cluster in Fleet.clusters {
#Platform: Components: "\(Cluster.name)/{{ .Name }}": {
path: "components/cert-manager"
cluster: Cluster.name
}
}
}

View File

@@ -0,0 +1,16 @@
package holos
// Produce a helm chart build plan.
(#Helm & Chart).BuildPlan
let Chart = {
Name: "{{ .Name }}"
Version: #CertManager.Version
Namespace: #CertManager.Namespace
Repo: name: "{{ .RepoName }}"
Repo: url: "{{ .RepoURL }}"
Values: installCRDs: true
Values: startupapicheck: enabled: false
}

View File

@@ -0,0 +1,10 @@
{
"name": "cert-manager",
"short": "cloud native X.509 certificate management for kubernetes",
"long": "cert-manager creates tls certificates for workloads in your kubernetes cluster and renews the certificates before they expire.",
"chart": "",
"reponame": "jetstack",
"repourl": "https://charts.jetstack.io",
"version": "1.15.3",
"namespace": "cert-manager"
}

View File

@@ -0,0 +1,15 @@
package holos
// Manage the component on every cluster in the platform
for Fleet in #Fleets {
for Cluster in Fleet.clusters {
#Platform: Components: "\(Cluster.name)/external-secrets-crds": {
path: "projects/platform/components/external-secrets-crds"
cluster: Cluster.name
}
#Platform: Components: "\(Cluster.name)/external-secrets": {
path: "projects/platform/components/external-secrets"
cluster: Cluster.name
}
}
}

View File

@@ -0,0 +1,10 @@
package holos
// Platform wide configuration
#ExternalSecrets: {
Version: "{{ .Version }}"
Namespace: "external-secrets"
}
// Register the namespace
#Namespaces: (#ExternalSecrets.Namespace): _

View File

@@ -0,0 +1,33 @@
package holos
import (
"encoding/yaml"
ks "sigs.k8s.io/kustomize/api/types"
)
(#Kubernetes & {Name: "external-secrets-crds"}).BuildPlan
// Holos stages BuildPlan resources as an intermediate step of the rendering
// pipeline. The purpose is to provide the resources to kustomize for
// post-processing.
let BuildPlanResources = "build-plan-resources.yaml"
_Kustomization: ks.#Kustomization & {
apiVersion: "kustomize.config.k8s.io/v1beta1"
kind: "Kustomization"
resources: [
// Kustomize the intermediate build plan resources.
BuildPlanResources,
// Mix-in external resources.
"https://raw.githubusercontent.com/external-secrets/external-secrets/v\(#ExternalSecrets.Version)/deploy/crds/bundle.yaml",
]
}
// Generate a kustomization.yaml directly from CUE so we can provide the correct
// version.
spec: components: kubernetesObjectsList: [{
// intermediate build plan resources to kustomize. Necessary to activate the
// kustomization post-rendering step in holos.
kustomize: resourcesFile: BuildPlanResources
kustomize: kustomizeFiles: "kustomization.yaml": yaml.Marshal(_Kustomization)
}]

Some files were not shown because too many files have changed in this diff Show More