[SERVER-80977] Abort chunk migration after cloning more data than expected Created: 12/Sep/23 Updated: 26/Oct/23 |
|
| Status: | Blocked |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Sergi Mateo Bellido | Assignee: | Backlog - Catalog and Routing |
| Resolution: | Unresolved | Votes: | 0 |
| Labels: | oldshardingemea | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Assigned Teams: |
Catalog and Routing
|
||||||||
| Operating System: | ALL | ||||||||
| Participants: | |||||||||
| Description |
|
In the cloning phase, we should define some kind of limit (probably based on the max chunk size) that once it's reached migrations are aborted. This scenario might happen if the chunk being migrated is also receiving a lot of data or if the balancer wrongly decided to move a chunk whose real size was much bigger than the approximated size. Note that the approximated size of a chunk is computed as NumberOfDocsInTheChunk * averageSizePerDoc. Note that those chunks should be marked as jumbo, otherwise the balancer will deterministically try to move them again. |