[dev-v2.10] rancher-vsphere-csi 105.0.1+up3.3.1-rancher7 Add option priority class for csi node daemonset

[epacke]

Checkpoints for Chart Bumps

release.yaml:

• Each chart version in release.yaml DOES NOT modify an already released chart. If so, stop and modify the versions so that it releases a net-new chart.
• Each chart version in release.yaml IS exactly 1 more patch or minor version than the last released chart version. If not, stop and modify the versions so that it releases a net-new chart.

Chart.yaml and index.yaml:

• The index.yaml file has an entry for your new chart version.
• The index.yaml entries for each chart matches the Chart.yaml for each chart.
• Each chart has ALL required annotations
  • kube-version annotation
  • rancher-version annotation
  • permits-os annotation (indicates Windows and/or Linux)

---

Issue:

Nodes under memory pressure may evict the csi-node daemonset pods. Which in turn prevents pods depending on PVCs from starting.

Solution

My suggestion is to add a priority class option to the helm chart

QA Testing Considerations

Induce memory pressure on a node by creating a pod which uses upp all memory available for pods.

[github-actions]

Validation steps

• Ensure all container images have repository and tag on the same level to ensure that all container images are included in rancher-images.txt which are used by airgap customers.

  Ex:-
    longhorn-controller:
      repository: rancher/hardened-sriov-cni
      tag: v2.6.3-build20230913
  

• Add a 👍 (thumbs up) reaction to this comment once done. CI won't pass without this reaction to the github-action bot's latest validation comment.
• Approve the PR to run the CI check.