-
Type: Task
-
Resolution: Duplicate
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: Sharding
-
Sharding NYC
-
3
The coordinator observer will see the write to config.reshardingOperations and see that the write changes a shard to error. The coordinator will then update itself to the error state. Then it will refresh all shards. After all shards have returned that they've updated to the error state, the coordinator should indefinitely attempt to remove all resharding metadata that currently exists.
Also take care of all TODOS linked in SERVER-53006 in this ticket, then close both in one commit. These TODOS are currently tied to the duplicated ticket SERVER-51800.
- duplicates
-
SERVER-54000 Make errors propagate from the ReshardingCoordinator to participants
- Closed
- has to be finished together with
-
SERVER-53006 Complete TODO listed in SERVER-51800
- Closed
- is duplicated by
-
SERVER-51800 Refresh both donor and recipient shards in case of resharding error state
- Closed