XDCR Error in 6.5.0 CE edition

Couchbase Server
1

XDCR stuck between two Ubuntu 16 clusters in 6.5.0 after replicating 50% documents, Some errors and warnings in Log

Source number of nodes: 2
Target number of nodes: 3
Is mismatch between no of nodes can be an issue?

This is only one way replicator.

2020-04-28 13:04:01 10.139.100.176:genericPipeline.context.Start:Failed to start service CheckpointManager for 66e0a79048d1b73458ceded33a2c2de7/XXX err=MCResponse status=AUTH_ERROR, opcode=SASL_AUTH, opaque=0, msg: {“error”:{“ref”:“f692d276-f6ae-4333-f004-eee361c3f828”}}

2020-04-28T13:06:13.402+05:30 WARN GOXDCR.CheckpointMgr: Failed to construct memcached client for ** custom.domain.com** :11210, err=MCResponse status=AUTH_ERROR, opcode=SASL_AUTH, opaque=0, msg: {“error”:{“ref”:“0d508c49-463a-4116-b736-09aec1dedc98”}}

2020-04-28T13:10:28.589+05:30 ERRO GOXDCR.CapiSvc: Calling _pre_replicate on custom.domain.com :8091 failed for vb=122, err=Specified resource does not exist.

2020-04-28T13:08:11.686+05:30 ERRO GOXDCR.CapiSvc: Received error when making rest call or unmarshalling data. baseURL= ** custom.domain.com** :8091, path=_goxdcr/_pre_replicate/bucket?bucket_uuid=195e02e48bc59a08276d837de6b94967, err=Specified resource does not exist., statusCode=401, num_retry=4

around 40% documents got transferred in all 3 destination nodes however it is now stuck. Pause and resume dint work.

I also followed instructions mentioned in below forum but still no luck.

Can any one please help?

2

It got resolved, Putting my scenario for others to benefit from

It seems that one of the node in source cluster was using private ip.
I thought for sourcer -> destination XDCR we only need to open communication from source to target ip
But it seems that one of the node in source cluster was using private ip so target nodes cant connect and were throwing above errors.
As a workaround I created another cluster in source datacenter which has the private ip visibility and made suare the ip was public for that new cluster. from that new cluster i transferred docs to target. That was one time task so it was fine for me.