-
Notifications
You must be signed in to change notification settings - Fork 1.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
teleport-cluster chart: set strategy to 'Recreate' if chartMode is 'standalone' (fixes #11484) #11493
teleport-cluster chart: set strategy to 'Recreate' if chartMode is 'standalone' (fixes #11484) #11493
Conversation
Hey @johanneskastl - thanks for the PR. Please take a look at https://github.com/gravitational/teleport/blob/master/examples/chart/CONTRIBUTING.md for some additional testing we'd need. |
cb61471
to
35af5b8
Compare
Thanks for the hint @webvictim! I have added a test to Did I miss any other steps? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@johanneskastl Thanks! Looks great!
Just one final ask: could you also please add three extra tests to make sure that the strategy is not set (so will default to RollingUpdate
) for the other modes (aws
, gcp
and custom
)?
Something similar to this section: https://github.com/gravitational/teleport/blob/master/examples/chart/teleport-cluster/tests/deployment_test.yaml#L234-L310 - you should be able to use the same values files.
35af5b8
to
f828785
Compare
Hey @webvictim, I added three more tests for AWS/GCP/Custom mode to not contain the Should I also commit this file that was created by running
|
14f97df
to
8a2df42
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @johanneskastl,
that looks good for me! Looking forward for this PR as it blocks us too currently.
8a2df42
to
49fb486
Compare
Looks good to me @johanneskastl - we just need to commit the Thanks for the contribution! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. As @zmb3 says, please commit the snapshot and we'll be good to go!
/gcbrun |
81dcc3a
to
f1b293d
Compare
@webvictim @zmb3 I commited the file, please have a look. |
23b1138
to
2f7c4b9
Compare
f0fa0c5
to
de1ec3c
Compare
Rebased on current master and solved conflict in Can this be merged? Or is something missing? |
/gcbrun |
…gy to 'Recreate' if chartMode is 'standalone' (fixes gravitational#11484) Signed-off-by: Johannes Kastl <[email protected]>
…for strategy in standalone mode Signed-off-by: Johannes Kastl <[email protected]>
…test.yaml.snap after running local tests
de1ec3c
to
db1640d
Compare
@webvictim I'll stop rebasing this, as I cannot run the tests anyway. Please ping me if I can assists or if fixes are needed. |
@johanneskastl Thanks - I'm sure it's fine, we'll take a look when people are back at work on Monday. |
/gcbrun |
1 similar comment
/gcbrun |
It looks like the review bot is refusing to approve this because reasons. I'll open a new buddy PR and get it merged. |
Thanks @webvictim |
* examples/chart/teleport-cluster/templates/deployment.yaml: set strategy to 'Recreate' if chartMode is 'standalone' (fixes #11484) Signed-off-by: Johannes Kastl <[email protected]> * examples/chart/teleport-cluster/tests/deployment_test.yaml: add test for strategy in standalone mode Signed-off-by: Johannes Kastl <[email protected]> * update examples/chart/teleport-cluster/tests/__snapshot__/deployment_test.yaml.snap after running local tests Co-authored-by: Johannes Kastl <[email protected]>
@webvictim I see you merged 11693 containing my changes, so can we close this one? |
@johanneskastl Thanks for the contribution! |
Pull request was closed
* examples/chart/teleport-cluster/templates/deployment.yaml: set strategy to 'Recreate' if chartMode is 'standalone' (fixes #11484) Signed-off-by: Johannes Kastl <[email protected]> * examples/chart/teleport-cluster/tests/deployment_test.yaml: add test for strategy in standalone mode Signed-off-by: Johannes Kastl <[email protected]> * update examples/chart/teleport-cluster/tests/__snapshot__/deployment_test.yaml.snap after running local tests Co-authored-by: Johannes Kastl <[email protected]>
* helm: Buddy merge for #11493 (#11693) * examples/chart/teleport-cluster/templates/deployment.yaml: set strategy to 'Recreate' if chartMode is 'standalone' (fixes #11484) Signed-off-by: Johannes Kastl <[email protected]> * examples/chart/teleport-cluster/tests/deployment_test.yaml: add test for strategy in standalone mode Signed-off-by: Johannes Kastl <[email protected]> * update examples/chart/teleport-cluster/tests/__snapshot__/deployment_test.yaml.snap after running local tests Co-authored-by: Johannes Kastl <[email protected]> * helm: Allow probe timeouts to be configurable (buddy merge of #11176) (#11396) * Allow for probe timeouts to be configurable When setting up a new Teleport enterprise cluster on GCP, I noticed that I needed to set the probe timeouts to get the cluster to be healthy. This seems to be a known issue (kubernetes/kubernetes#89898). As a "stopgap", I've updated the helm chart to allow for end users to be able to configure these timeouts. * Update configuration option name and add documentation * Update docs/pages/kubernetes-access/helm/reference.mdx Co-authored-by: Gus Luxton <[email protected]> * Add tests for probeTimeoutSeconds * Add probeTimeoutSeconds to required values * Add probeTimeoutSeconds to teleport-kube-agent * Add tests for probeTimeoutSeconds to teleport-kube-agent * Add probeTimeoutSeconds to teleport-kube-agent reference Co-authored-by: Hunter Madison <[email protected]> Co-authored-by: Hunter Madison <[email protected]> Co-authored-by: Johannes Kastl <[email protected]> Co-authored-by: Hunter Madison <[email protected]> Co-authored-by: Hunter Madison <[email protected]>
Fixes #11484
In standalone mode a PVC is created. The volume cannot be attached to more than one pod (depending on the storage available), so using a
Recreate
strategy is needed to avoid stuck deployments on upgrades.