[SERVER-33729] Integration test for multi-statement transaction involving multiple read operations and no writes Created: 07/Mar/18 Updated: 03/May/18 Resolved: 03/May/18 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Replication, Storage |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | David Storch | Assignee: | Tess Avitabile (Inactive) |
| Resolution: | Duplicate | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Sprint: | Repl 2018-05-21 | ||||||||
| Participants: | |||||||||
| Description |
|
We should write a test verifying that a transaction can consist of multiple read commands at readConcern level snapshot, not just a single read command, or a find/aggregate and its subsequent getMores. All reads in the transaction should see the same snapshot, and should be isolated from concurrent writes. |
| Comments |
| Comment by Tess Avitabile (Inactive) [ 03/May/18 ] |
|
Yes, I agree this will be covered by |
| Comment by William Schultz (Inactive) [ 30/Apr/18 ] |
|
I think this will end up being covered by |
| Comment by Eric Milkie [ 08/Mar/18 ] |
|
I would expect the Replica Set Transactions project would add testing for all combinations of reads and writes. A multi-statement transaction that happens to do no writes is just a specialization of multi-statement transactions. |
| Comment by David Storch [ 07/Mar/18 ] |
|
spencer, this seems to fall somewhere in between the local snapshot reads project and the multi-document transactions project. For now, I've put it into local snapshot reads. Let me know if this is wrong, and your team in fact plans to do this work. It looks like this functionality has already been implemented, but we don't have the tests to prove it. |