i have below setup
source : 1 cbnode 1 syncgateway
destination : 2 cbnodes in cluster 1 syncgateway
sg-replicate for replication for replication between source syncgateway to destination syncgateway
As per my understanding ,sg-replicate creates sync:local reference document(which hold the sequence number of source syncgateway) in destination cluster for future replication purpose. it is working fine with happy scenarios.
whenever one of couchbase nodes in destination cluster goes down and sync:local reference document exists in other active nodes, in this case sg-replicate will get chekpoint sequence number and some documents are replicated to active node by sg-replicate and some documents are not replicated which are mapped to down node. and also sync:local reference document updated with new sequence number.
In above some documents are missed to replicate to destination cluster. we can only get those documents with altering sync:local reference document.
it is something we can stop replication till all destination couchbase nodes are up and running.
has anyone experienced this scenario.