[SERVER-45901] Make moveChunk robust to being killOp'd after commit has been sent to the config but before the node has found out the commit decision Created: 31/Jan/20 Updated: 29/Oct/23 Resolved: 05/Feb/20 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Sharding |
| Affects Version/s: | None |
| Fix Version/s: | 4.3.4 |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | Esha Maharishi (Inactive) | Assignee: | Esha Maharishi (Inactive) |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Backwards Compatibility: | Fully Compatible |
| Sprint: | Sharding 2020-02-10 |
| Participants: |
| Description |
|
Currently, if the MigrationSourceManager's thread is killOp'd after commit has been sent to the config but before this node has found out the decision, then the range deletion tasks and migration coordinator doc won't be updated. This is because _cleanup won't set an abort decision on the MigrationCoordinator, and _cleanup will call MigrationCoordinator::completeMigration without a decision having been set, which will just log an error and return without doing anything. |
| Comments |
| Comment by Githook User [ 05/Feb/20 ] |
|
Author: {'name': 'Esha Maharishi', 'username': 'EshaMaharishi', 'email': 'esha.maharishi@mongodb.com'}Message: |