[SERVER-79298] FCBIS should succeed in unhealthy set Created: 25/Jul/23  Updated: 31/Jul/23

Status: Open
Project: Core Server
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Improvement Priority: Major - P3
Reporter: Matthew Russotto Assignee: Backlog - Replication Team
Resolution: Unresolved Votes: 0
Labels: former-quick-wins
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
Assigned Teams:
Replication
Participants:

 Description   

File Copy Based Initial Sync requires the majority point of the set to advance in order to start an extension round. This means if the set is unhealthy the sync will not succeed. We should fix this, either by

1) If extension fails before we've copied anything, logging an error and continuing with the sync. This may leave the node more lagged than desired, but it will catch up with normal replication (and perhaps make the set healthy again, if this was a re-sync)

and/or

2) Remove the requirement of majority point advance when extended the backup cursor. To get a valid extension, it is only necessary that the no-holes point be advanced beyond the extension time, not the majority. However, doing this may affect backup which may be expecting extensions to be majority committed, so we should check this; if that is the case we will need to parameterize .


Generated at Thu Feb 08 06:40:37 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.