-
Type: Task
-
Resolution: Won't Do
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
Cluster Scalability
We typically allow a transaction participant shard to retry on stale errors (StaleConfig, StaleDbVersion, ShardCannotRefreshDueToLocksHeld) if the request started the transaction on the shard, but we disallow retrying at all if the participant shard acted as a sub-router. We can consider allowing a shard that acted as a sub-router to retry if it targeted 0 or 1 other participant shards.