server: make visible the ranges that fail to move during a decommission #76249
Labels
A-cli-server
CLI commands that pertain to CockroachDB server processes
A-kv-decom-rolling-restart
Decommission and Rolling Restarts
A-kv-observability
C-enhancement
Solution expected to add code/behavior + preserve backward-compat (pg compat issues are exception)
T-server-and-security
DB Server & Security
When a replica transfer stalls during a decommission, we currently have no visibility of which replicas are “stuck.” The only progress indicator is a counter of the remaining replicas on the node being decommissioned, without any indications for what ranges these replicas are for. Today, discovering these ranges requires a manual search.
We need visibility of the ranges of the stalling replicas in order to optimize our ability diagnose stuck decommissioning processes.
Needed for #74158.
Epic CRDB-11843
Jira issue: CRDB-13046
The text was updated successfully, but these errors were encountered: