We have a 3 node cluster and we have a performance test being run with 2500 ops per second, does this create 2500 ops times 3 nodes since it writes to all the 3 nodes (copies of the same document)
It will only perform the specified number of operations from client -> cluster nodes.
However depending on how many replicas you have configured, there will be additional mutations replicated between the nodes - for each replica there’s an additional copy of the data.