From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: MIME-Version: 1.0 References: <20181002124329.21248-1-linus.walleij@linaro.org> <05fdbe23-ec01-895f-e67e-abff85c1ece2@kernel.dk> <1538582091.205649.20.camel@acm.org> In-Reply-To: From: Paolo Valente Date: Wed, 3 Oct 2018 18:09:58 +0200 Message-ID: Subject: Re: [PATCH] block: BFQ default for single queue devices To: Bart Van Assche Cc: Jens Axboe , Linus Walleij , linux-block , linux-mmc , linux-mtd@lists.infradead.org, Pavel Machek , Ulf Hansson , Richard Weinberger , Artem Bityutskiy , Adrian Hunter , Jan Kara , Andreas Herrmann , Mel Gorman , Chunyan Zhang , Linux-Kernal , "'Paolo Valente' via bfq-iosched" , Oleksandr Natalenko , Mark Brown Content-Type: multipart/alternative; boundary="000000000000c1947705775545db" List-ID: --000000000000c1947705775545db Content-Type: text/plain; charset="UTF-8" Il mer 3 ott 2018, 18:02 Paolo Valente ha scritto: > > > > Il giorno 03 ott 2018, alle ore 17:54, Bart Van Assche < > bvanassche@acm.org> ha scritto: > > > > On Wed, 2018-10-03 at 08:29 +0200, Paolo Valente wrote: > >> [1] https://lkml.org/lkml/2017/2/21/791 > >> [2] http://algo.ing.unimo.it/people/paolo/disk_sched/results.php > >> [3] https://lwn.net/Articles/763603/ > > > > From [2]: "BFQ loses about 18% with only random readers, because the > number > > of IOPS becomes so high that the execution time and parallel efficiency > of > > the schedulers becomes relevant." Since the number of I/O patterns for > which > > results are available on [2] is limited and since the number of devices > for > > which test results are available on [2] is limited (e.g. RAID is > missing), > > there might be other cases in which configuring BFQ as the default would > > introduce a regression. > > > > From [3]: none with throttling loses 80% of the throughput when used > to control I/O. > Of course, the same happens with mq-deadline plus throttling, sorry. Paolo > > --000000000000c1947705775545db Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable


= Il mer 3 ott 2018, 18:02 Paolo Valente <paolo.valente@linaro.org> ha scritto:


> Il giorno 03 ott 2018, alle ore 17:54, Bart Van Assche <bvanassche@= acm.org> ha scritto:
>
> On Wed, 2018-10-03 at 08:29 +0200, Paolo Valente wrote:
>> [1] https://lkml.org/lkml/2017/2/21/791 >> [2] http://algo.ing.= unimo.it/people/paolo/disk_sched/results.php
>> [3] https://lwn.net/Articles/763603/
>
> From [2]: "BFQ loses about 18% with only random readers, because = the number
> of IOPS becomes so high that the execution time and parallel efficienc= y of
> the schedulers becomes relevant." Since the number of I/O pattern= s for which
> results are available on [2] is limited and since the number of device= s for
> which test results are available on [2] is limited (e.g. RAID is missi= ng),
> there might be other cases in which configuring BFQ as the default wou= ld
> introduce a regression.
>

>>From [3]: none with throttling loses 80% of the throughput when used
to control I/O.

Of course, the same happens with mq-deadline plus throttling= , sorry.

Paolo

--000000000000c1947705775545db--