A recent escalation, HELP-92299, showed that customers can hit failures during manual chunk-management workflows and receive error output that is too generic to diagnose or recover from easily, creating manual overhead and reducing confidence in operating the cluster. In that escalation, the customer was running manual split and moveChunk operations, hit precondition/concurrency-related failures, and needed guidance on how to understand and avoid those conflicts.

This ticket should review the error messages returned by split and mergeChunks, especially for precondition and concurrency-related failures, and improve them so they explain the likely cause, whether the condition is transient, and what the operator should check or do next. The goal is to make these failures self-manageable without requiring an escalation, for example by pointing users toward conflicts with concurrent chunk operations or relevant cluster state checks when applicable.