Information about the Cluster
Rancher Server Setup

• Rancher version: v2.7.4 upgraded to v2.7-head 0bcf068 rolled back to 2.7.2
• Installation option (Docker install/Helm Chart): Helm Chart
  • If Helm Chart, Kubernetes Cluster and version (RKE1, RKE2, k3s, EKS, etc): RKE1, v1.25.6, v1.4.5 Hardened
• Proxy/Cert Details: Self-signed

Information about the Cluster

• Kubernetes version: 1.25 RKE2 and RKE1 node driver clusters
• Cluster Type (Local/Downstream): 1 etcd, 1 cp, 1 worker

Describe the bug
When we upgrade and roll back rancher server from v2.7.4 >> v2.7-head >> v2.7.4 or v2.7.2 >> v2.7.3 >> v2.7.2 , downstream clusters webhook version is not upgraded but the local cluster webhook version is downgraded to the required version.

To Reproduce

1. Install rancher on version 2.7.4
2. Create a custom RKE1 and RKE2 cluster with the specs mentioned above.
3. Install rancher backup/restore chart version in the local cluster with config set to S3
4. Take a backup from the rancher backup/restore page and the backupfile gets saved in the s3.
5. Upgrade rancher server to v2.7-head
6. Rollback to 2.7.4 by following the docs
7. Bring down the nodes of current rancher local cluster
8. Migrate rancher to v2.7.0 by following instructions here https://ranchermanager.docs.rancher.com/v2.7/how-to-guides/new-user-guides/backup-restore-and-disaster-recovery/restore-rancher-launched-kubernetes-clusters-from-backup

Result
Once the migration is successful, and rancher is installed if we navigate to downstream clusters, we notice the webhook version stays at the version of v2.7-head and is not compatible with the local cluster webhook version.

Expected Result
Webhook on downstream clusters stay compatible with local cluster