Several roles and scripts in the osh-infra repo have defaults for
ubuntu that are still on xenial when we mostly run bionic or newer.
This change updates these references to default to focal.
Change-Id: I84bca2d685cf1c67b10eee3bb7c05deb2cc4670b
This change updates many of the deployment scripts to properly
handle deploying each service via helm 3 and updates each job
to use the helm v3 install script.
Change-Id: I90a7b59231376b9179439c2554e46449d59b9c15
With the move to helm v3, helm status requires a namespace to be specified, but doing so breaks helm v2 compatability. This change removes the usage of helm serve in openstack-helm-infra's deployment scripts.
Change-Id: I8e035d70dd652d5253f534ad6b28042347158ff4
This removes the functionality to perform envsubst in the feature
gate script to prevent users with specific env set running into
unexpected error. This feature will be re-visited in the future to
be made more robust.
Signed-off-by: Tin Lam <tin@irrational.io>
Change-Id: I6dcfd4dad138573294a9222e4e7af80c9bff4ac0
About zookeeper chart,It's been removed,But there are still some related scripts that have not been completely deleted,we should remove them.
Change-Id: Iae20717482ad6c7a40f54174eef120d094abbd59
The deal is that all the jobs' scripts include extra arguments
when deploying helm charts, except these ones in the commit.
It would be useful to use override files in these charts.
+ Fix typo in apparmor.yaml for node-exporter
+ Amend apparmor.yaml for openstack-exporter since
those values are already by default in values.yaml
Change-Id: Ibe8b38977216e618dccba7e8443b3cc05a772de5
New systemd 237-3ubuntu10.43 bumps memlock limit from 16 to 64 MB [1]
which seems to cause issues with eBPF related operations in containers
run with root [2] as a possible root cause.
Here we have an option to downgrade systemd to previous available
version or to set previous default memlock limit to systemd defaults or
docker unit. Setting systemd DefaultLimitMEMLOCK in this commit.
[1] https://launchpad.net/ubuntu/+source/systemd/237-3ubuntu10.43
[2] https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1837580/comments/9
Change-Id: I55d14ffa47a7a29d059f2f3b502bb38be0a5dd3d
Signed-off-by: Andrii Ostapenko <andrii.ostapenko@att.com>
Switch from using images from defined in docker_images to provided as
zuul artifacts. Currently to be used in conjunction with [0] in
openstack-helm-images pipelines.
[0] https://review.opendev.org/741551
Change-Id: I43dbd38906e8854c87a361f2e5e479f57850252f
Signed-off-by: Andrii Ostapenko <andrii.ostapenko@att.com>
If the test pod still exists, then the new test run
fails with ERROR: pods "abc-test" already exists
So, Removing remaining test pods before new test run
Change-Id: I3b3ed5ceaf420aa39a669b4a50a838ad154b1fdd
Closes-Bug: #1882030
1) A separate prometheus job need to provide target and scrap metrics
2) it is based on https://github.com/prometheus/blackbox_exporter
Adding script file for deployment and job under zuul.d
Resolving conflict
Change-Id: Ia15ab7d8ef882886fe0e37cc2599e6815d7bcc6c
With this commit minikube is installed using contents of precreated
minikube-aio image containing installation script, all required binaries
and images inside. Pulling a single image from dockerhub via opendev
dockerhub proxy and loading images allows to save up to 6 minutes in
minikube installation.
Change-Id: I5936f440eb0567b8dcba2fdae614e4c5e88a7b9a
Signed-off-by: Andrii Ostapenko <andrii.ostapenko@att.com>
This chart could deploy fluentd either as a Deployment
or a Daemonset. Both options would use the deployment-fluentd
template with various sections toggled off based on values.yaml
I'd like to know - Does anyone run this chart as a Deployment?
We can simplify the chart, and zuul gates, by changing the chart
to deploy a Daemonset specifically.
Change-Id: Ie88ceadbf5113fc60e5bb0ddef09e18fe07a192c
This is to update ceph scripts to create loopback devices
in single script and also to update gate scripts.
Change-Id: Id6e3c09dca20d98fcbcc434e65f790c06b6272e8
Along with fixing the bug, with this we'll decrease build time for
all jobs using minikube and collecting tiller metrics for more than
2 minutes.
Change-Id: Ia166584eae48c643248f977b959aa6336e3a327e
Signed-off-by: Andrii Ostapenko <andrii.ostapenko@att.com>
change docker image to point to the latest metacontroller image.
change python image to point to version 3.7
add updateStrategy to CompositeController.
add replicas config to DaemonJobController via zuul gate.
Change-Id: I2a48bc6472017802267980fe474d81886113fcda
- This is to make use of loopback devices for ceph osds since
support for directory backed osds going to deprecate.
- Move to bluestore from filestore for ceph-osds.
- Seperate DB and WAL partitions from data so that gates will validate
the scenario where we will have fast storage disk for DB and WAL.
Change-Id: Ief6de17c53d6cb57ef604895fdc66dc6c604fd89
This adds a chart for the node problem detector. This chart
will help provide additional insight into the status of the
underlying infrastructure of a deployment.
Updated the chart with new yamllint checks.
Change-Id: I21a24b67b121388107b20ab38ac7703c7a33f1c1
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This change updates the fluentd configmap-etc to render
.Values.conf.fluentd.template as a template, allowing for greater
flexibility in configuration.
Change-Id: I8809767c679c377e319ecc53960c55ae18e1b558
The current copyright refers to a non-existent group
"openstack helm authors" with often out-of-date references that
are confusing when adding a new file to the repo.
This change removes all references to this copyright by the
non-existent group and any blank lines underneath.
Change-Id: I1882738cf9757c5350a8533876fd37b5920b5235
This PS updates the bind mounts for ceph logs directorys to be
emptydirs. This ensures we do not polute the hosts permanantly
with ceph logs, which should be directed to stdout.
Change-Id: I6d72c0864b9ecc493cd62564e0e0450d90cfcf00
Signed-off-by: Pete Birley <pete@port.direct>
This change updates the fluentd chart to use the native
fluent-plugin-prometheus for metric production. This plugin
provides more detailed metrics about fluentd's operations,
specifically regarding input and output statistics.
https://github.com/fluent/fluent-plugin-prometheus
Using the plugin, each fluentd pod produces metrics, so
scape annotations have been added to the pod spec. The
zuul check on metric producers has been updated to account
for this.
Depends-On: https://review.opendev.org/714167
Change-Id: I809356d92b0cff1e31cb2062102bbedefd4843fd
XPath to expected element was changed after
Kibana upgrade, this commit changes XPath
according chnges in new Kibana.
Change-Id: I501de225e1226991db9c263cedf38397cda7b51f
These charts bootstraps a metacontroller on a
Kubernetes cluster using the Helm package manager.
This enables you to deploy custom controllers as
service + deployment pairs. A DaemonJobController
chart bootstraps the CompositeController and register
DaemonJob CRD, the daemonjob controller executes
DaemonJob(CR's) in kubernetes Cluster.
Change-Id: Ic946f564ea1cf07e89c90a598e59230dc240950c
This updates the overrides provided for deploying fluentd as a
daemonset to get kernel messages from the journal instead of
/var/log/kern.log directly, and also uses the journal to get
messages associated with logging to auth.log (syslog facility
10). This provides additional metadata and
a cleaner interface for gathering these logs via fluentd
Change-Id: I8e832db276095771d6a869e998d7a69795dfee37
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This change updates the Ceph charts to use Ceph Nautilus images
built on Ubuntu Bionic instead of Xenial. The mirror that hosts
Ceph packages only provides Nautilus packages for Bionic at
present, so this is necessary for Nautilus deployment.
There are also several configuration and scripting changes
included to provide compatibility with Ceph Nautilus. Most of
these simply allow existing logic to execute for Nautilus
deployments, but some logical changes are required to support
Nautilus as well.
NOTE: The cephfs test has been disabled because it was failing
the gate. This test has passed in multiple dev environments, and
since cephfs isn't used by any openstack-helm-infra components we
don't want this to block getting this change merged. The gate
issue will be investigated and addressed in a subsequent patch
set.
Change-Id: Id2d9d7b35d4dc66e93a0aacc9ea514e85ae13467
This updates the osh-infra-logging single node job to omit the
fluentbit deployment step, as having multiple logging daemonsets
deployed to the single node jobs is causing IO issues. Also, it
was noted that the fluentd-deployment step was missing the
overrides to move the fluentd-deployment release from utilizing a
daemonset to a deployment. This resulted in 3 logging daemons
being deployed to a single host
Change-Id: I4a0c5550e6ea6a331aab0082a975f161e65704bf
Signed-off-by: Steve Wilkerson <sw5822@att.com>
