Hi
Is it possible to setup an XDCR cross data center replication with less number of nodes on the secondary DC cluster. In other words for a 8 node cluster (4 data + 4 index/query) can be replicated to a 2 node cluster (1 data + 1 index/query) on other DC is there any minimal limit of nodes based on type in setting up XDCR replication
we are using replica config as 1 at bucket level
thanks
Yes – you can setup an XDCR (cross data center replication) between clusters with different number of nodes.
XDCR replication works between two independent clusters – the cluster topology (number of nodes, number of replicas, services, etc) does not have to match. You can create a replication for a bucket or a single collection, so you’ll just need to make sure that the destination (target) cluster has adequate resources for the initial replication of data, and then, can keep up with the changes (mutations) from the source.
Note: You can add/remove nodes from both the source and the target clusters while the XDCR replication is on-going, and XDCR will automatically handle the cluster topology changes. Part of the XDCR internal handling of cluster topology changes is to wait until the changes stabilize before continuing replication, so there may be a short slow down of replications – in Server 7.6, XDCR introduced a “[Source or Target]: Topology change detected” informational message for awareness.