On Wed, 2018-10-03 at 08:29 +-0200, Paolo Valente wrote: +AD4 +AFs-1+AF0 https://lkml.org/lkml/2017/2/21/791 +AD4 +AFs-2+AF0 http://algo.ing.unimo.it/people/paolo/disk+AF8-sched/results.php +AD4 +AFs-3+AF0 https://lwn.net/Articles/763603/ >From +AFs-2+AF0: +ACI-BFQ loses about 18+ACU with only random readers, because the number of IOPS becomes so high that the execution time and parallel efficiency of the schedulers becomes relevant.+ACI Since the number of I/O patterns for which results are available on +AFs-2+AF0 is limited and since the number of devices for which test results are available on +AFs-2+AF0 is limited (e.g. RAID is missing), there might be other cases in which configuring BFQ as the default would introduce a regression. I agree with Jens that it's best to leave it to the Linux distributors to select a default I/O scheduler. Bart.
