I have single cluster with two nodes. Both are active and Intra-Cluster Replication enabled.
I tried to test INSERT and GET operations using curl & query service.
When one of the nodes is down, all INSERT operations are getting failed mentioning that node is not reachable (“code”:12009).
My understanding is during Intra-Cluster Replication, when one node is down the DB opertaions are automatically carried out by other node which is alive. However this is not happenning.
Is there any way to acheive this?
Hi @Dinakar -
There’s a lot to digest here - Intra-Cluster Replication | Couchbase Docs
Every document is assigned to an active vbucket; vbuckets are on nodes. If the active vbucket for a document is on a node that is down, the document cannot be written. For the replica vbucket to become the active vbucket, a failover is required. Failover | Couchbase Docs. Once the failover has taken place, all the replica vbuckets on the remaining node will have changed to be active vbuckets and the client can get a new vbucket-to-node map and will be able to insert once again.
Ok, understood. Thanks for the response.