From: Peter Zijlstra <peterz@infradead.org>
To: Huaixin Chang <changhuaixin@linux.alibaba.com>
Cc: linux-kernel@vger.kernel.org, bsegall@google.com,
dietmar.eggemann@arm.com, juri.lelli@redhat.com, mgorman@suse.de,
mingo@redhat.com, pauld@redhead.com, pjt@google.com,
rostedt@goodmis.org, vincent.guittot@linaro.org,
khlebnikov@yandex-team.ru, xiyou.wangcong@gmail.com,
shanpeic@linux.alibaba.com
Subject: Re: [PATCH 1/4] sched/fair: Introduce primitives for CFS bandwidth burst
Date: Thu, 17 Dec 2020 14:36:56 +0100 [thread overview]
Message-ID: <20201217133656.GX3040@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20201217074620.58338-2-changhuaixin@linux.alibaba.com>
On Thu, Dec 17, 2020 at 03:46:17PM +0800, Huaixin Chang wrote:
> In this patch, we introduce the notion of CFS bandwidth burst. Unused
> "quota" from pervious "periods" might be accumulated and used in the
> following "periods". The maximum amount of accumulated bandwidth is
> bounded by "burst". And the maximun amount of CPU a group can consume in
> a given period is "buffer" which is equivalent to "quota" + "burst in
> case that this group has done enough accumulation.
Oh man, Juri, wasn't there a paper about statistical bandwidth
accounting somewhere? Where, if you replace every utilization by a
statistical variable, the end result is still useful?
That is, instead of something like; \Sum u_i <= 1, you get something
like: \Sum {avg(u),var(u)}_i <= {1, sqrt(\Sum var_i^2)} and you can
still proof bounded tardiness etc.. (assuming a gaussian distribution).
The proposed seems close to that, but not quite, and I'm afraid it's not
quite strong enough to still provide any guarantees.
next prev parent reply other threads:[~2020-12-17 13:38 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-17 7:46 [PATCH 0/4] sched/fair: Burstable CFS bandwidth controller Huaixin Chang
2020-12-17 7:46 ` [PATCH 1/4] sched/fair: Introduce primitives for CFS bandwidth burst Huaixin Chang
2020-12-17 13:36 ` Peter Zijlstra [this message]
2020-12-21 13:53 ` changhuaixin
2021-01-12 9:21 ` changhuaixin
2020-12-17 7:46 ` [PATCH 2/4] sched/fair: Make CFS bandwidth controller burstable Huaixin Chang
2020-12-18 9:53 ` kernel test robot
2020-12-17 7:46 ` [PATCH 3/4] sched/fair: Add cfs bandwidth burst statistics Huaixin Chang
2020-12-17 7:46 ` [PATCH 4/4] sched/fair: Add document for burstable CFS bandwidth control Huaixin Chang
2020-12-17 21:25 ` [PATCH 0/4] sched/fair: Burstable CFS bandwidth controller Benjamin Segall
2021-01-20 12:27 ` [PATCH v2 " Huaixin Chang
2021-01-20 12:27 ` [PATCH 1/4] sched/fair: Introduce primitives for CFS bandwidth burst Huaixin Chang
2021-01-20 12:27 ` [PATCH 2/4] sched/fair: Make CFS bandwidth controller burstable Huaixin Chang
2021-01-20 17:06 ` kernel test robot
2021-01-20 18:33 ` kernel test robot
2021-01-20 22:01 ` kernel test robot
2021-01-20 22:01 ` [RFC PATCH] sched/fair: __refill_cfs_bandwidth_runtime() can be static kernel test robot
2021-01-20 12:27 ` [PATCH 3/4] sched/fair: Add cfs bandwidth burst statistics Huaixin Chang
2021-01-20 12:27 ` [PATCH 4/4] sched/fair: Add document for burstable CFS bandwidth control Huaixin Chang
2021-01-20 19:48 ` Randy Dunlap
2021-01-21 11:04 ` [PATCH v3 0/4] sched/fair: Burstable CFS bandwidth controller Huaixin Chang
2021-01-21 11:04 ` [PATCH v3 1/4] sched/fair: Introduce primitives for CFS bandwidth burst Huaixin Chang
2021-03-10 12:39 ` Peter Zijlstra
2021-01-21 11:04 ` [PATCH v3 2/4] sched/fair: Make CFS bandwidth controller burstable Huaixin Chang
2021-03-10 13:04 ` Peter Zijlstra
2021-03-12 13:54 ` changhuaixin
2021-03-16 10:55 ` Peter Zijlstra
2021-01-21 11:04 ` [PATCH v3 3/4] sched/fair: Add cfs bandwidth burst statistics Huaixin Chang
2021-03-10 13:10 ` Peter Zijlstra
2021-01-21 11:04 ` [PATCH v3 4/4] sched/fair: Add document for burstable CFS bandwidth control Huaixin Chang
2021-03-10 13:17 ` Peter Zijlstra
2021-01-26 10:18 ` [PATCH v3 0/4] sched/fair: Burstable CFS bandwidth controller changhuaixin
2021-03-10 12:26 ` Peter Zijlstra
2021-02-09 13:17 ` Odin Ugedal
2021-02-09 10:28 ` Tejun Heo
2021-02-27 13:48 ` changhuaixin
2021-03-10 11:11 ` Odin Ugedal
2021-03-12 13:26 ` changhuaixin
-- strict thread matches above, loose matches on Subject: below --
2021-02-02 11:40 [PATCH " Huaixin Chang
2021-02-02 11:40 ` [PATCH 1/4] sched/fair: Introduce primitives for CFS bandwidth burst Huaixin Chang
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20201217133656.GX3040@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=bsegall@google.com \
--cc=changhuaixin@linux.alibaba.com \
--cc=dietmar.eggemann@arm.com \
--cc=juri.lelli@redhat.com \
--cc=khlebnikov@yandex-team.ru \
--cc=linux-kernel@vger.kernel.org \
--cc=mgorman@suse.de \
--cc=mingo@redhat.com \
--cc=pauld@redhead.com \
--cc=pjt@google.com \
--cc=rostedt@goodmis.org \
--cc=shanpeic@linux.alibaba.com \
--cc=vincent.guittot@linaro.org \
--cc=xiyou.wangcong@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox