Details
-
Improvement
-
Resolution: Fixed
-
Major
-
2.8.0
-
Security Level: Public
-
CBG Sprint 65
-
1
Description
Once the cfgEventCallback fires for "_sync:cfg" prefixed docs, I'm not sure any of the normal DocChanged processing needs to take place for these documents.
By not returning early, we're falling through to normal processing, and ending up with warnings in the logs:
2020-10-21T11:53:15.863+01:00 [INF] SGCluster: Replication definitions changed - refreshing...
|
2020-10-21T11:53:15.863+01:00 [DBG] SGCluster+: c:sg_int_1-cfgSG cfg_sg: Get, key: sgrCluster, cas: 0
|
2020-10-21T11:53:15.863+01:00 [DBG] CRUD+: No xattr content found for key=_sync:cfgsgrCluster, xattrKey=_sync: sub-document path does not exist
|
2020-10-21T11:53:15.864+01:00 [DBG] Replicate+: Aligning state for existing replication repl1
|
2020-10-21T11:53:15.864+01:00 [WRN] changeCache: Doc "_sync:cfgsgrCluster" does not have valid sync data. -- db.(*changeCache).DocChanged() at change_cache.go:450
|
Attachments
Issue Links
- Clones
-
CBG-1151 Exit early from DocChange after running callback for _sync:cfg docs
-
- Closed
-