Running CachyOS With The BORE Scheduler While Disabling Ananicy-CPP

Source Domain: www.phoronix.com

Last week I ran benchmarks of CachyOS with the BORE scheduler using its “linux-cachyos-bore” kernel option. The results didn’t end up being as enticing as anticipated but the developer behind the BORE scheduler commented in the forums that he recently received reports from users experiencing game stuttering while using BORE that was attributed to CachyOS’ default use of Ananicy-Cpp. So over the weekend I did another CachyOS BORE run without that CachyOS default.

Ananicy-cpp is a system utility that tries to optimize Linux performance by dtnamically managing process priorities. Those wanting to learn more about it can do so via this GitHub repository for the CachyOS rules applied.

With Ananicy-cpp being used by default on CachyOS, I ran another linux-cachyos-bore run but with disabling Ananicy-cpp that can be easily done via the CachyOS Hello app.

So here’s a look at those additional CachyOS benchmark figures:

While disabling Ananicy-CPP with BORE is reported to help reduce game stuttering issues, in most of the benchmarks I ran there wasn’t any real difference with Ananicy-CPP disabled.

One area observed where disabling Ananicy-CPP did help the linux-cachyos-bore performance was in FIO’s sequential write benchmarks.

Disabling it was a mixed bag for the Ethr network local host benchmark runs.

Those interested can see all the conducted benchmarks via this result file, but long story short besides the noted benefits in FIO IO_uring for sequential writes, disabling Ananicy-CPP with BORE didn’t provide much of an observable benefit in these particular benchmarks carried out.

Source