fix: pods rotation

[rugggger]

TL;DR

Introduces cluster configuration hash tracking to trigger pod rotation when cluster spec changes beyond just image updates, enabling proper rolling updates for configuration changes.

What changed?

• Added PodConfigVersion environment variable configuration and WekaRuntimeVersion constant to track runtime changes
• Implemented clusterConfigHash() function that generates SHA-256 hash from cluster image, version numbers, and tracked spec fields (network, tolerations, hugepages, etc.)
• Modified upgrade controller to accept target config hash alongside target image, with new logic to check both image and config alignment
• Added TargetClusterSpecHash field to container specs and ClusterSpecHashAnnotation to pod annotations for tracking
• Updated container reconciliation to trigger pod rotation when either image or cluster config hash changes
• Refactored spec propagation into reusable ApplyUpdatableSpecToContainer() and PropagateSpecToContainer() functions
• Enhanced upgrade flows to handle both image upgrades and configuration-only changes

How to test?

1. Deploy a WekaCluster and verify initial pod creation includes config hash annotations
2. Change cluster spec fields like network configuration, tolerations, or hugepages settings without changing image
3. Verify containers get updated with new TargetClusterSpecHash and pods are rotated to apply changes
4. Test image-only upgrades still work as expected
5. Test combined image + config changes trigger proper rolling updates

Why make this change?

Previously, only image changes would trigger pod rotation during upgrades. Configuration changes to cluster specs (network settings, resource allocations, tolerations, etc.) would update container specs but not rotate pods, leaving running pods with stale configurations. This change ensures any meaningful cluster configuration change triggers proper pod rotation to apply the new settings.

[github-actions]

Dependency Review

✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.

Snapshot Warnings

⚠️: No snapshots were found for the head SHA 7fb8bbd.

Ensure that dependencies are being submitted on PR branches and consider enabling retry-on-snapshot-warnings. See the documentation for more information and troubleshooting advice.

Scanned Files

None

[rugggger]

• fix: propagation for untracked spec fields #2404
• fix: pods rotation #2394 👈 (View in Graphite)
• main

---

How to use the Graphite Merge Queue

Add the label main-merge-queue to this PR to add it to the merge queue.

You must have a Graphite account in order to use the merge queue. Sign up using this link.

An organization admin has required the Graphite Merge Queue in this repository.

Please do not merge from GitHub as this will restart CI on PRs being processed by the merge queue.

This stack of pull requests is managed by Graphite. Learn more about stacking.

[graphite-app]

Graphite Automations

"Add anton/matt/sergey/kristina as reviwers on operator PRs" took an action on this PR • (03/25/26)

3 reviewers were added to this PR based on Anton Bykov's automation.

[kristina-solovyova]

what if ComputeExtraCores is changed, but we are drive wekacontainer?
will it mean that hash is chnaged then and pod should be rotated?

[rugggger]

yes, all tracked fields will cause a rotation similar to an upgrade (for all containers, same order, one by one, but no image-specific operations)

[kristina-solovyova]

it seems like this hash should be calculated per role so that we do not rotate pods that won't actually get any updates

[rugggger]

yeah, sounds right