ACK: [SRU][K/L][PATCH 0/1] fix cgroups v1 blkio.throttle statistics
Tim Gardner
tim.gardner at canonical.com
Tue May 9 12:58:04 UTC 2023
On 5/8/23 11:18 PM, Andrea Righi wrote:
> BugLink: https://bugs.launchpad.net/bugs/2016186
>
> [Impact]
>
> Commit f382fb0bcef4 ("block: remove legacy IO schedulers") introduced a
> behavior change in the blkio throttle cgroup subsystem: IO statistics
> are not reported anymore unless a throttling rule is explicitly defined,
> because the current code only counts bios that are actually throttled.
>
> This behavior change is potentially breaking some user-space
> applications that are relying on the old behavior (see original bug
> report).
>
> [Test case]
>
> - mount cgroup v1
> - create a blkio cgroup
> - move a task into the blkio cgroup
> - perform some I/O (i.e., dd)
> - read the IO stats for the cgroup (blkio.throttle.io_serviced and
> blkio.throttle.io_service_bytes in cgroupfs)
> - IO stats are all 0, unless a throttle rule is defined
>
> Previous behavior (kernel 5.15) was showing I/O statistics even without
> throttling rules defined.
>
> [Fix]
>
> Apply / backport this fix:
> https://lore.kernel.org/lkml/20230507170631.89607-1-hanjinke.666@bytedance.com/t/
>
> [Regression potential]
>
> The fix is affecting the block IO cgroup subsystem, we may see potential
> regressions in this particular cgroup subsystem with this fix applied.
>
> [Additional notes]
>
> This fix is provided as SAUCE patch for now, but it is likely to be
> applied to the upstream stable trees at this point (it has been
> submitted upstream and well tested).
>
>
Acked-by: Tim Gardner <tim.gardner at canonical.com>
--
-----------
Tim Gardner
Canonical, Inc
More information about the kernel-team
mailing list