Details
-
Bug
-
Resolution: Fixed
-
Major
-
4.5.0
-
Untriaged
-
Unknown
Description
When one of the upstream DCP connection, for the same feed, is closed it leads to panic in worker.go and which should be handled by feed by cleaning up the bucket book-keeping.
WORKAROUND: We can turn off multiple DCP stream perf optimization (no major impact doing so) to avoid this bug
curl -X POST -u <user:pass> http://localhost:9102/settings --data '{"projector.dcp.numConnections": 1}'
|
Attachments
Issue Links
For Gerrit Dashboard: MB-20142 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
65722,2 | MB-20142: fix worker pool panic and cleanupbucket(). | unstable | indexing | Status: ABANDONED | 0 | 0 |
66095,6 | MB-20142: fix worker pool panic and cleanupbucket(). | unstable | indexing | Status: MERGED | +2 | +1 |