My current use case of rpc-perf with explicitly sized dataset involves first configuring rpc-perf to send write heavy traffic to warm up the cache, then restarting rpc-perf with the workload being tested. It would be nice if we had a "warm up" option that would systematically write all keys, then automatically transition to the configured workload.
My current use case of rpc-perf with explicitly sized dataset involves first configuring rpc-perf to send write heavy traffic to warm up the cache, then restarting rpc-perf with the workload being tested. It would be nice if we had a "warm up" option that would systematically write all keys, then automatically transition to the configured workload.