I have restored cassandra snapshot in one node in a new cluster which consists of 6 nodes. Will it replicate to the remaining nodes? Or do we need to copy the snapshot to remaining nodes (as done in node1)?
We have created keyspace and tables in the node1 it is replicated in all the nodes, but the snapshot data is not replicated.
Ok, I think I see the question:
will it replicate to remaining nodes or need to copy the snapshot to remaining nodes as done in node1?
So, replication only happens from write activity. As you're restoring a snapshot, replication isn't going to work automatically. Yes, you will need to copy the snapshot files to the remaining nodes and run the same procedure.
Once the nodes have all joined, be sure to run a nodetool cleanup on each node. This will delete any data which is present on other nodes, but that the particular node is not responsible for.
When restoring snapshots on a node, the data will only exist on that node and will not be replicated to the replicas in the cluster.
Also, depending on the configuration of the source and the destination cluster, the procedure you described may not be valid.
If the destination cluster has an identical configuration to the source cluster (same token assignments, same topology), then you will need to clone the corresponding snapshots to the rest of the nodes in the cluster. For step-by-step details, see the procedure I documented in How do I restore Cassandra snapshots to another cluster with identical configuration?.
If the two clusters are not identical, copying a snapshot from the source is not going to work because the data it contains won't necessarily belong to the destination node. In this scenario, you can clone the data to the new cluster with the bulk-load method using sstableloader. For steps on how to do this, see the procedure I documented in How do I migrate data in tables to a new Cassandra cluster?. Cheers!
