Support etcd snapshot create + restore (Imported cluster day 2 ops)

[eva-vashkevich]

• for imported and capi clusters add snapshots tab to detail page
  • list v1/rke.cattle.io.etcdsnapshots (with same columns as v2prov)
    • create snapshot via new operation
      • visible if user can create the operation CR
    • delete snapshot as before
      • visible if user can delete the etcd snapshot resource
    • restore snapshot via new operation
      • visible if user can create the operation CR
• cluster actions
  • take snapshot via new operation
    • visible if user can create the operation CR
  • restore snapshot - model updated with new operation
    • visible if user can create the operation CR
• snapshot configuration
  • TBD automatic snapshots, snapshot location, metrics

rancher/rancher#54259
rancher/rancher#54263

[github-actions]

🎯 Agentic Issue Triage

Enhancement request to extend etcd snapshot create and restore functionality to imported RKE2/K3s and CAPI clusters in the Rancher Dashboard. This is a sub-task of #16453 "Day 2 Ops for Imported RKE2/K3S Clusters" and is classified as Priority 1.

📎 Related Issues

Issue	Title	Notes
#16453	Day 2 Ops for Imported RKE2/K3S Clusters	Parent umbrella issue (priority/0)
#16135	Limit restore mode options in UI based on ETCDSnapshot annotation	Related
#16136	Snapshot restore state reused when creating on demand snapshot	Related
#16202	Display error when a user attempts RKE2/K3s disaster recovery restore from local snapshot	Related
#14632	[BUG] etcd snapshot on S3 - Incorrect Age	Related

🔬 Technical Analysis

Type: Feature Enhancement

Scope of work:

1. Add snapshots tab to imported and CAPI cluster detail pages (listing v1/rke.cattle.io.etcdsnapshots)
2. Create snapshot via new operation CR (RBAC: can create operation CR)
3. Delete snapshot (RBAC: can delete etcd snapshot resource)
4. Restore snapshot via new operation CR (RBAC: can create operation CR)
5. Cluster-level actions: take/restore snapshot via new operation
6. Snapshot configuration (automatic snapshots, snapshot location, metrics — TBD)

Key files to reference:

• shell/models/rke.cattle.io.etcdsnapshot.js — existing model (currently only supports restore for v2prov)
• shell/detail/provisioning.cattle.io.cluster.vue — has showSnapshots() computed, rke2Snapshots, snapshot headers, but showSnapshots() returns false for non-RKE2 clusters
• shell/components/PromptRestore.vue — existing restore dialog
• shell/edit/provisioning.cattle.io.cluster/tabs/etcd/index.vue — existing etcd tab

Key technical decisions:

• The showSnapshots() computed in shell/detail/provisioning.cattle.io.cluster.vue must be extended for imported and CAPI clusters
• Need to implement new operation CRs for create/restore (different from the existing rkeConfig.etcdSnapshotCreate/etcdSnapshotRestore approach used for v2prov)
• RBAC checks must differ: create/restore use "can create operation CR", delete uses "can delete etcd snapshot resource"
• Need to determine how imported clusters expose v1/rke.cattle.io.etcdsnapshots

✅ Sub-Tasks Checklist

• Extend snapshots tab to imported/CAPI cluster detail pages
• List v1/rke.cattle.io.etcdsnapshots (same columns as v2prov)
• Add "Create Snapshot" action via new operation CR with RBAC guard
• Add "Delete Snapshot" action with RBAC guard
• Add "Restore Snapshot" action via new operation CR with RBAC guard
• Add cluster-level "Take Snapshot" action via new operation
• Update cluster-level "Restore" modal for new operation
• Determine TBD snapshot configuration (automatic, location, metrics)

🛠️ Debugging & Investigation Tips

• Check what operation CRs are available for imported clusters (vs v2prov's rkeConfig.etcdSnapshotCreate/Restore)
• Test with an actual imported RKE2/K3s cluster and CAPI cluster to verify API endpoint availability
• Check the RFD linked in [EPIC] Day 2 Ops for Imported RKE2/K3S Clusters #16453: https://github.com/SUSE/rancher-architecture/pull/38
• Start from showSnapshots() in shell/detail/provisioning.cattle.io.cluster.vue and trace what conditions need to change for imported/CAPI clusters

Generated by Agentic Triage for issue #17263 · ◷

[github-actions]

🤖 Issue grooming analysis

This feature request is well-groomed and ready for triage. Here's the analysis against the checklist:

✅ Clear Problem Statement
The title and description clearly identify the need: add etcd snapshot create and restore capabilities as day-2 operations for imported and CAPI clusters.

✅ Acceptance Criteria
A detailed, testable checklist is provided covering: snapshots tab on the detail page (listing v1/rke.cattle.io.etcdsnapshots), create/delete/restore snapshot operations, cluster-level actions (take snapshot, restore snapshot), and visibility conditions based on user permissions to create/delete the relevant operation CRs.

✅ Technical Details
Specific API resources (v1/rke.cattle.io.etcdsnapshots) are named, the feature scope (imported clusters, CAPI clusters, v2prov parity) is defined, and conditions for action visibility (based on ability to create/delete operation CRs) are described.

✅ Reproducibility (Feature)
A clear feature flow is described: from the cluster detail page snapshots tab through to cluster-level snapshot actions, covering all CRUD operations.

This analysis was generated automatically based on the repository's grooming checklist.

Generated by Daily Issue Grooming · ◷