[SERVER-20875] Update or remove connPoolSync command Created: 12/Oct/15 Updated: 06/Dec/22 |
|
| Status: | Backlog |
| Project: | Core Server |
| Component/s: | Networking |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Improvement | Priority: | Major - P3 |
| Reporter: | Samantha Ritter (Inactive) | Assignee: | Backlog - Service Architecture |
| Resolution: | Unresolved | Votes: | 0 |
| Labels: | sa-remove-fv-backlog-22 | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Assigned Teams: |
Service Arch
|
||||||||
| Participants: | |||||||||
| Description |
|
connPoolSync is an undocumented internal command that flushes the global pool, deleting all of its pooled connections. Given that the server now contains many connection pools in addition to the global client connection pool, this command no longer works the way it claims to. We should either update this command to do something better, or remove it entirely. |
| Comments |
| Comment by Randolph Tan [ 23/Nov/15 ] |
|
The machinery for making gle work used to be stored in thread local variables per connection. The connPoolSync was meant to make sure that the test starts with a clean slate. We also asked users to call connPoolSync to get around some connection related issues in the past. |
| Comment by Andy Schwerin [ 23/Nov/15 ] |
|
Supposing that the only use of this command is in unit tests, the only relevant users are the following:
renctan wrote create_index_gle.js. I don't know why it calls connPoolSync. sharding_state_after_stepdown.js has a comment that implies it's intentionally trying to flush the global connection pool, but I'm not certain why. The auth unit tests are just confirming that correct access control behavior for the connPoolSync command, so should not count as uses. |