Redis 4.x Cookbook

There's more...

It is critical to note that the benchmarking result just gives you an extreme testing result for your Redis alone and you should never use it as a stress test for your whole application's performance. Moreover, if you want to simulate the real application environment, you should start the Redis benchmark tool on the same host that the applications get deployed.

In this way, the overhead of network hop is the same for both the benchmark tool and the applications. It can be a big help for you to get a more practical testing result. Lastly, if you want to do a benchmark of a Redis Cluster with N shards, the performance of the Cluster can be roughly estimated through the benchmarking result against one of the N shards multiplied ...