Details
-
Bug
-
Resolution: Fixed
-
Critical
-
2.5.0
-
Security Level: Public
-
Triaged
-
Centos 64-bit
Description
Dest:
http://172.23.96.11:8091/
Source:
http://172.23.96.15:8091/
Initial replication of 2 source buckets overloaded destination causing well-known ep-engine deadlock. Eventually Erlang crashes started to happen on source side.
Live cluster is available, will collect/share diags and dumps.
Attachments
Issue Links
- relates to
-
MB-9612 xdcr spams all cluster management error logs on tmp errors on destination cluster
- Closed
For Gerrit Dashboard: MB-10127 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
33302,1 | MB-10127 Drop messages in error_logger if mailbox is to large. | master | ns_server | Status: ABANDONED | 0 | +1 |
33335,3 | MB-10127 Drop messages in error_logger if mailbox is to large. | for-rackaware | ns_server | Status: MERGED | +2 | +1 |
33416,1 | Merge remote-tracking branch 'origin/for-rackaware' | master | ns_server | Status: MERGED | +2 | +1 |
33421,3 | MB-10127 Get rid of enable_pipeline_ops setting. | for-rackaware | ns_server | Status: MERGED | +2 | +1 |
33422,3 | MB-10127 Always use memcached connection pool for xdcr. | for-rackaware | ns_server | Status: MERGED | +2 | +1 |
33423,3 | MB-10127 Remove dead code. | for-rackaware | ns_server | Status: MERGED | +2 | +1 |
33424,3 | MB-10127 Eliminate xdc_vbucket_rep_xmem_{srv,worker} modules. | for-rackaware | ns_server | Status: MERGED | +2 | +1 |
33590,1 | Merge remote-tracking branch 'origin/for-rackaware' | master | ns_server | Status: MERGED | +2 | +1 |