I have a 4-node CB cluster with version 4.5.1 (recently updated from 4.1.0) and a backup server with version 4.1.0. In the backup datacenter there was a network failure. When the problem was removed, XDCR automatically resumed on 3 nodes but on one it fails to start. In the log I have the following notice:
PipelineManager 2017-08-29T22:44:28.570+02:00 [INFO] Replication Status = map[06189a4a72b9f0553d754f56502b5e0a/default/savecart_bkp:name={06189a4a72b9f0553d754f56502b5e0a/default/saveca
rt_bkp}, status={Pending}, errors={[{“time”:“2017-08-29T20:00:35.573676789+02:00”,“errMsg”:“Pipeline 06189a4a72b9f0553d754f56502b5e0a/default/savecart_bkp failed to start, err=map[xmem_
06189a4a72b9f0553d754f56502b5e0a/default/savecart_bkp_cbbackup.savecart:11210_0:dial tcp xxx.xxx.xxx.xxx:11210: i/o timeout]\n”},{“time”:“2017-08-29T20:00:15.515165076+02:00”,“errMsg”:“Pip
eline 06189a4a72b9f0553d754f56502b5e0a/default/savecart_bkp failed to start, err=map[xmem_06189a4a72b9f0553d754f56502b5e0a/default/savecart_bkp_cbbackup.savecart:11210_0:dial tcp xxx.xxx.
xxx.xxx:11210: i/o timeout]\n”},{“time”:“2017-08-29T19:59:55.61649265+02:00”,“errMsg”:“Pipeline 06189a4a72b9f0553d754f56502b5e0a/default/savecart_bkp failed to start, err=map[xmem_06189a
4a72b9f0553d754f56502b5e0a/default/savecart_bkp_cbbackup.savecart:11210_0:dial tcp xxx.xxx.xxx.xxx:11210: i/o timeout]\n”},{“time”:“2017-08-29T19:59:30.8380718+02:00”,“errMsg”:“dcp_06189a4
a72b9f0553d754f56502b5e0a/default/savecart_bkp_cb4.savecart:11210_1:DCP stream has been closed.”}]}, progress={The runtime context has been started}
I tried to pause and resume XDCR process but nothing helps.
Does anyone has any idea what might be a problem?