[SERVER-37847] Refactor cluster_aggregate to use one collation Created: 31/Oct/18 Updated: 27/Oct/23 Resolved: 28/Aug/20 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Querying |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Improvement | Priority: | Major - P3 |
| Reporter: | Ian Boros | Assignee: | Nikita Lapkov (Inactive) |
| Resolution: | Gone away | Votes: | 0 |
| Labels: | neweng, qexec-team | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Sprint: | Query 2020-09-07 |
| Participants: |
| Description |
|
In a code review, Charlie pointed out that in dispatchShardPipeline() we sometimes pass the collation from the request AggregationRequest when doing shard targeting, rather than passing the collation which is the default for the collection (which is computed here). We looked into it some more and could not find any incorrect behavior, but it might be worth doing a more thorough check, and refactor the code so that we only use one collation. |
| Comments |
| Comment by Nikita Lapkov (Inactive) [ 28/Aug/20 ] |
|
nicholas.zolnierz confirmed that the problem is gone |
| Comment by Ian Boros [ 27/Aug/20 ] |
|
nikita.lapkov Hello! Indeed it looks like this has been fixed. I think we can close this as "Gone Away." IIRC Nick Zolnierz knows this code pretty well (and is the author of the relevant code in master), so I would just check with him before closing it. |