[SERVER-85088] POC: Explore ways we can shrink the generated oplog entries from pipeline-style updates Created: 01/Apr/19 Updated: 12/Jan/24 Resolved: 18/Dec/19 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | Charlie Swanson | Assignee: | Charlie Swanson |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Sprint: | Query 2019-06-17, Query 2019-12-02, Query 2019-12-16, Query 2019-12-30 | ||||||||
| Participants: | |||||||||
| Description |
|
The existing update system has a fairly sophisticated way of tracking damages performed by an update to generate a minimal oplog entry. We have a fairly sophisticated system within aggregation to detect which fields are modified by which stages - we may be able to re-use it to inform oplog entry generation. |
| Comments |
| Comment by Charlie Swanson [ 18/Dec/19 ] |
|
Just chatted with ian.boros who's going to be working on the design. We have a rough POC here which doesn't handle $unsets which could be interesting and different, but we don't think there's remaining value in this ticket, so we'll call it complete. |
| Comment by Craig Homa [ 12/Dec/19 ] |
|
Hey charlie.swanson, is there a link to the POC document that you can share? |
| Comment by Charlie Swanson [ 19/Jun/19 ] |
|
We're going to need to scope this project (PM-1460) before getting started meaningfully so I'm punting this ticket out of the sprint to be revisited after scoping. |
| Comment by Charlie Swanson [ 30/May/19 ] |
|
We have some initial perf results to consider. I'll gather them and report here. |