Re: [PATCH v2] sched: async unthrottling for cfs bandwidth

[Tejun Heo]

On Wed, Nov 02, 2022 at 06:10:49PM +0100, Michal Koutný wrote:
> On Tue, Nov 01, 2022 at 12:38:23PM -1000, Tejun Heo <tj@xxxxxxxxxx> wrote:
> > > We're in the process of transitioning to using bw instead for this
> > > instead in order to maintain parallelism. Fixing bw is definitely
> > > going to be useful, but I'm afraid we'll still likely have some issues
> > > from low throughput for non-bw reasons (some of which we can't
> > > directly control, since arbitrary jobs can spin up and configure their
> > > hierarchy/threads in antagonistic ways, in effect pushing out the
> > > latency of some of their threads).
> > 
> > Yeah, thanks for the explanation. Making the lock more granular is tedious
> > but definitely doable. I don't think I can work on it in the near future but
> > will keep it on mind. If anyone's interested in attacking it, please be my
> > guest.
> 
> From my experience, throttling while holding kernel locks (not just
> cgroup_mutex) causes more trouble than plain cgroup_mutex scalability
> currently.

Oh yeah, absolutely. Low cpu bw config + any shared kernel resource is a
nightmare and this thread was originally about addressing that.

Thanks.

-- 
tejun