[docs] Establish Cadence For Benchmarking Against Competitors

[mischif]

Dragonfly claims a significant performance increase over Redis, but the numbers appear out of date - as best I can tell from the README, Dragonfly last benchmarked against Redis in December of 2023 but Redis 8 was released in the last few weeks with claims of significant performance improvements.

Additionally, new competitors such as Valkey have entered the space but Dragonfly appears not to have benchmarked against them at all.

I am requesting that Dragonfly rerun its benchmarks against a recent release of both Redis and Valkey, while noting the versions used, and create a process to update the benchmarks every year or so.

I can help create some automation around running the benchmarks if necessary.

[vyavdoshenko]

Hi @mischif Thanks for reporting this issue. Have you tried to read the Dragonfly engineering blog section? Engineering Posts This blog constantly exposes different aspects of Dragonfly and its competitors' performance, throughput, latency, and so on. For instance, I found the latest one from 25 March, 2025: Dragonfly vs. Valkey Benchmark: 4.5x Higher Throughput on Google Cloud

[mischif]

I had not looked at the engineering blog, it's great to see a recent comparison between Dragonfly and Valkey (that post also reminded me this would've been better suited as a github discussion).

I'm assuming the tests used to benchmark Valkey in that post would also work for Redis as well? If so, is there anything preventing someone from wrapping them up with some setup code to create a simple benchmark script that could be used to update the readme? I don't mind doing it if there's consensus on how to benchmark the servers correctly.