I have been testing the throughput of CockroachDB with the tool TPC-C.
When testing with different number of nodes and the same workload, the throughput stays basically the same. I had expected it to increase as the number of nodes increased. Why is this?
Example of result:
Nr of nodes: 3, then 6, then 9
Throughput: 3669, 3658, 3673
Efficiency (how close the system is to its theoretical maximum: 95.1, 94.8, 95.2
Latency: 61, 61, 49