-
Notifications
You must be signed in to change notification settings - Fork 7
refactor: Tidy up Nutanix CSI with consistent apply strategy #733
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
Conversation
32f4ff4
to
50f485d
Compare
c91fd70
to
df37085
Compare
50f485d
to
8fa5cf2
Compare
hack/addons/kustomize/nutanix-snapshot-csi/kustomization.yaml.tmpl
Outdated
Show resolved
Hide resolved
df37085
to
71b92d2
Compare
e5717e3
to
e31bb3b
Compare
71b92d2
to
1315c65
Compare
e31bb3b
to
a2332de
Compare
1315c65
to
5869295
Compare
a2332de
to
13fdac2
Compare
5869295
to
f2628f9
Compare
13fdac2
to
ced4eb6
Compare
f2628f9
to
68020de
Compare
This PR/issue depends on:
|
Move the current logic out to a helm addon strategy as done in other providers. Also move Helm release to kube-system to be consistent with other CSI providers. And tidy up flag names to use `--csi.nutanix.` prefix for flags, including wiring up to Helm chart values and deployment. Finally, add e2e test for CSI deployment in Nutanix infra.
The Helm chart does not propagate the required env/arg with the namespace, and it does not support setting extra args/envs.
ced4eb6
to
ee8e7af
Compare
This is not true anymore, right? |
Thanks! Removed from PR description which will be used as commit message on merge. |
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.
Thanks for cleaning this up.
nit: I think the e2e changes could have been a separate PR?
In future I would hope that changes like this would include e2e changes to improve the situation and validate the fixes actually work without someone running them manually. |
Currently both Nutanix CSI and AWS EBS CSI handlers deploy the snapshot-controller. Generally the snapshot-controller should be treated as a separate addon that all compatible CSI providers can take advantage of. This commit removes the snapshot-controller from both Nutanix and AWS EBS CSI, and instead deploys the snapshot-controller separately so they can both use it as necessary. Deployment tested manually on AWS via e2e test suite as well as the automated Docker test suite running in CI. Depends on #733.
🤖 I have created a release *beep* *boop* --- ## 0.11.0 (2024-06-27) <!-- Release notes generated using configuration in .github/release.yaml at main --> ## What's Changed ### Exciting New Features 🎉 * feat: Configure namespace sync in helm chart by @dlipovetsky in #726 * feat: Support CRS for local-path provisioner and add CSI e2e by @jimmidyson in #737 * feat: Support HelmAddon strategy for AWS EBS by @jimmidyson in #732 * feat: Deploy snapshot-controller as separate addon by @jimmidyson in #734 * feat: Update AWS CCM versions and add HelmAddon strategy by @jimmidyson in #748 ### Fixes 🔧 * fix: Namespace Sync controller should list no resources when source namespace is empty string by @dlipovetsky in #725 * fix: Temporarily hard-code supported PC version for Nutanix CSI by @jimmidyson in #751 * fix: skip kubeadm CA file when Secret doesn't have a CA by @dkoshkin in #752 * fix: Correctly report failed deploy of ServiceLoadBalancer by @dlipovetsky in #759 ### Other Changes * build: Tidy up goreleaser config by @jimmidyson in #745 * ci: Fix up image loading for lint-test-helm by @jimmidyson in #746 * refactor: Tidy up Nutanix CSI with consistent apply strategy by @jimmidyson in #733 * test(e2e): Set empty env vars for Nutanix e2e vars by @jimmidyson in #749 * refactor: Use recommended "default" function syntax in helm templates by @dlipovetsky in #750 * refactor: Reusable HelmAddon strategy by @jimmidyson in #735 * test(e2e): Various e2e tests fixes by @jimmidyson in #754 * test(e2e): Correct default helm release names for AWS CCM and EBS CSI by @jimmidyson in #756 **Full Changelog**: v0.10.0...v0.11.0 --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Move the current logic out to a helm addon strategy as done in other providers.
And tidy up flag names to use
--csi.nutanix.
prefix for flags, including wiringup to Helm chart values and deployment.
Finally, add e2e test for CSI deployment in Nutanix infra.
Depends on #732.