The BatchWriteOp does not properly handle top level StaleConfig errors returned by the shard.
Â
I've discovered this as part of a refectoring attempt, in which I modified where we perform the collection acquisition by the shard during an insert operation, moving it some layers higher up. This change introduces the possibility of a StaleConfig exception being thrown much earlier in the write path on the shard (specifically, at this point).
- is depended on by
-
SERVER-108071 Handle top level StaleConfig errors in write commands
-
- Blocked
-
- is duplicated by
-
SERVER-115554 Timeseries insertions will fail in case of StaleDb for 8.0
-
- Closed
-
- related to
-
-
- Closed
-