[SERVER-26279] Sync source reevaluation can flap Created: 23/Sep/16 Updated: 06/Dec/22 Resolved: 04/Nov/16 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Replication |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Eric Milkie | Assignee: | Backlog - Replication Team |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||
| Assigned Teams: |
Replication
|
||||
| Operating System: | ALL | ||||
| Participants: | |||||
| Linked BF Score: | 0 | ||||
| Description |
|
The sync source reevaluation logic checks to see if any other sync source candidates are more than 30 seconds ahead of the current sync source. It uses heartbeat data to determine this, which means if no writes happen for longer than 30 seconds and then one write comes in, the different heartbeat timings may trigger the reevaluation unnecessarily. |
| Comments |
| Comment by Spencer Brody (Inactive) [ 04/Nov/16 ] |
|
This should be fixed by the introduction of the period noop writer that ensures that there is at least one write every 10 seconds. |