[SERVER-36538] Create idempotency tests for prepare, commit, and abort oplog entries Created: 08/Aug/18 Updated: 29/Oct/23 Resolved: 25/Apr/19 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Replication |
| Affects Version/s: | None |
| Fix Version/s: | 4.1.11, 4.1.13 |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | Judah Schvimer | Assignee: | Lingzhi Deng |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | prepare_testing | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||||||
| Backwards Compatibility: | Fully Compatible | ||||||||||||
| Sprint: | Repl 2019-04-22, Repl 2019-05-06 | ||||||||||||
| Participants: | |||||||||||||
| Comments |
| Comment by Githook User [ 29/May/19 ] |
|
Author: {'email': 'lingzhi.deng@mongodb.com', 'name': 'Lingzhi Deng', 'username': 'ldennis'}Message: |
| Comment by William Schultz (Inactive) [ 23/May/19 ] |
|
No problem, it should be easily fixed, but I just wanted to verify the intentions of the change, since we do utilize different SequenceType arguments in at least one test. |
| Comment by Lingzhi Deng [ 23/May/19 ] |
|
william.schultz, that was clearly a bug. I am sorry for my oversight. We should keep the prefixes/suffixes but instead calculate the prefixes/suffixes from the opPtrs. |
| Comment by William Schultz (Inactive) [ 23/May/19 ] |
|
lingzhi.deng It appears that this commit removed the logic inside IdempotencyTest::testOpsAreIdempotent for testing prefixes or suffixes of the given operation sequence. Was this intentional? |
| Comment by Githook User [ 25/Apr/19 ] |
|
Author: {'email': 'lingzhi.deng@mongodb.com', 'name': 'Lingzhi Deng', 'username': 'ldennis'}Message: |
| Comment by Max Hirschhorn [ 05/Feb/19 ] |
siyuan.zhou, to summarize what we discussed in person - the work May Hoque did back in the summer of 2017 was all built on top of the existing IdempotencyTest test fixture the Replication team had already created and is therefore based around oplog entries. |
| Comment by Siyuan Zhou [ 05/Feb/19 ] |
|
max.hirschhorn, does the idempontency test framework work at the granularity of commands or oplog entries? We used to assume one command is mapped to one oplog entry, but that's not true after larger transactions project. If the test framework works at command level, there might be extra work to do to extend this ticket for larger transactions. |