[SERVER-34150] Create a passthrough that does clean shutdowns Created: 27/Mar/18 Updated: 06/Dec/22 Resolved: 27/Feb/20 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Replication, Testing Infrastructure |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | Judah Schvimer | Assignee: | Backlog - Server Tooling and Methods (STM) (Inactive) |
| Resolution: | Won't Do | Votes: | 0 |
| Labels: | stm | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Assigned Teams: |
Server Tooling & Methods
|
||||||||
| Participants: | |||||||||
| Story Points: | 5 | ||||||||
| Description |
|
Recoverable rollback does work specifically to make fastcount correct across clean shutdown. A passthrough that does clean shutdowns on primaries and secondaries could catch some bugs here and around general data consistency. |
| Comments |
| Comment by Robert Guo (Inactive) [ 27/Feb/20 ] |
|
We have already a number of passthrough suites that do shutdowns; this request may no longer be necessary. |
| Comment by Spencer Brody (Inactive) [ 29/May/18 ] |
|
Moving this out of the Recoverable Rollback epic but leaving it open since it does still seem like there could be some value in this suite. But we don't think getting this coverage is necessary to release the new rollback implementation, since we believe we have sufficient coverage of all the pieces elsewhere. max.hirschhorn, if the TIG team isn't interested in doing this, feel free to close it. |
| Comment by Max Hirschhorn [ 05/Apr/18 ] |
|
The work on this ticket will involve creating a Python wrapper around the ReplSetTest#checkCollectionCounts() function in order to version that the WT size storer and a collection scan agree on the number of documents in the collection. |
| Comment by Judah Schvimer [ 02/Apr/18 ] |
|
The only thing this ticket adds on top of |
| Comment by Judah Schvimer [ 29/Mar/18 ] |
|
This appears to be covered by periodic_kill_secondaries, and we'll ensure the kill_primaries hook has the same behavior. Ensuring counts are maintained doesn't feel worth the engineering and compute cost. |