Redis Benchmark: Comparing throughput between data structures

So after running the internal Redis script for benchmarking, below are the results for throughput comparing Data Structures of String, List, Set, and Sorted Set.

First graph is for Writes:

Second graph is for Reads

Conclusion:

From the writes graph above, you can tell that sets perform very well when threads (concurrency) are only4 or less i.e.,when only 4 or less clients are hitting Redis at the same time, the String data structure has the best performance. But after 4 threads, you can see that is performs the worst of all the data structure. Coming off the block with 4 threads or less shows poor performance with Lists and Sorted Sets.

After 4 threads with writes, you can see that The List, Set, and Sorted Sets perform much better, almost equal. This means that with more clients hitting Redis at the same time, Sets, Lists and Sorted Sets perform much better in a multiple concurrent environment. With Reads, my conclusion that all data types are even matched. The difference is not too much, and hence results to choose which data stricture performs better are not conclusive.