From: Tao Ma <tm@tao.ma>
To: Vivek Goyal <vgoyal@redhat.com>
Cc: Shaohua Li <shli@kernel.org>, Tejun Heo <tj@kernel.org>,
axboe@kernel.dk, ctalbott@google.com, rni@google.com,
linux-kernel@vger.kernel.org, cgroups@vger.kernel.org,
containers@lists.linux-foundation.org
Subject: Re: IOPS based scheduler (Was: Re: [PATCH 18/21] blkcg: move blkio_group_conf->weight to cfq)
Date: Thu, 05 Apr 2012 00:06:14 +0800 [thread overview]
Message-ID: <4F7C7176.5090504@tao.ma> (raw)
In-Reply-To: <20120404151014.GD12676@redhat.com>
On 04/04/2012 11:10 PM, Vivek Goyal wrote:
> On Wed, Apr 04, 2012 at 07:52:24AM -0700, Shaohua Li wrote:
>
> [..]
>> Agreed, we can tweak cfq to make it support iops fairness because the two
>> are conceptually the same. The problem is if this is a mess. CFQ is quite
>> complicated already. In iops mode, a lot of code isn't required, like idle,
>> queue merging, thinktime/seek detection and so on, as the scheduler
>> will be only for ssd. With recent iocontext cleanup, the iops scheduler
>> code is quite short actually.
>
> Ok, this is somewhat a better reason to have a separate scheduler. I guess
> we need to look that actual iops code and that can help decide whether
> to keep it as a separate scheduler.
yes, actually I am afraid of making any *big* changes to cfq since it is
stable and complicated. It would be terrible for our customer which uses
sata and sas most of their time. So iops based scheduler is only used
for ssds.
>
> One question is still unanswered though. What real workload benefits
> from it? If you are not doing idling in iops based scheduler, I doubt
> you are doing to see much service differentiation on fast SSDs. As for
> service differentiation IO queues have to be continuously backlogged and
> total IOPS needed by applications need to be more than what disk can
> offer. It becomes very hard to produe continuously backlogged queues
> because real applications tend to read some data, proces data and then
> generate more IO.
OK, I guess I can describe our workload somehow. Yes, for very fast
SSDs, it would not help since the io depth is too high and we can't fill
in enough requests. But for some not that fast SSDs(say intel's x25m
series), it can only have tens of thousands iops, and it would help us
if we can have this type of ssd work as proportional. Yes, in most case,
the ssd will be idle, but we do have times that the disk is very busy
and we need the proportional iops to fit our customer's need.
Thanks
Tao
next prev parent reply other threads:[~2012-04-04 16:06 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-03-28 22:51 [PATCHSET] block: modularize blkcg config and stat file handling Tejun Heo
2012-03-28 22:51 ` [PATCH 01/21] blkcg: remove unused @pol and @plid parameters Tejun Heo
2012-03-28 22:51 ` [PATCH 02/21] blkcg: BLKIO_STAT_CPU_SECTORS doesn't have subcounters Tejun Heo
2012-03-28 22:51 ` [PATCH 03/21] blkcg: introduce blkg_stat and blkg_rwstat Tejun Heo
2012-03-28 22:51 ` [PATCH 04/21] blkcg: restructure statistics printing Tejun Heo
2012-03-28 22:51 ` [PATCH 05/21] blkcg: drop blkiocg_file_write_u64() Tejun Heo
2012-03-28 22:51 ` [PATCH 06/21] blkcg: restructure configuration printing Tejun Heo
2012-03-28 22:51 ` [PATCH 07/21] blkcg: restructure blkio_group configruation setting Tejun Heo
2012-03-28 22:51 ` [PATCH 08/21] blkcg: blkg_conf_prep() Tejun Heo
2012-03-28 22:53 ` Tejun Heo
2012-03-28 22:51 ` [PATCH 09/21] blkcg: export conf/stat helpers to prepare for reorganization Tejun Heo
2012-03-28 22:51 ` [PATCH 10/21] blkcg: implement blkio_policy_type->cftypes Tejun Heo
2012-03-28 22:51 ` [PATCH 11/21] blkcg: move conf/stat file handling code to policies Tejun Heo
2012-03-28 22:51 ` [PATCH 12/21] cfq: collapse cfq.h into cfq-iosched.c Tejun Heo
2012-03-28 22:51 ` [PATCH 13/21] blkcg: move statistics update code to policies Tejun Heo
2012-03-28 22:51 ` [PATCH 14/21] blkcg: cfq doesn't need per-cpu dispatch stats Tejun Heo
2012-03-28 22:51 ` [PATCH 15/21] blkcg: add blkio_policy_ops operations for exit and stat reset Tejun Heo
2012-03-28 22:51 ` [PATCH 16/21] blkcg: move blkio_group_stats to cfq-iosched.c Tejun Heo
2012-03-28 22:51 ` [PATCH 17/21] blkcg: move blkio_group_stats_cpu and friends to blk-throttle.c Tejun Heo
2012-03-28 22:51 ` [PATCH 18/21] blkcg: move blkio_group_conf->weight to cfq Tejun Heo
2012-04-01 21:09 ` Vivek Goyal
2012-04-01 21:22 ` Tejun Heo
2012-04-02 21:39 ` Tao Ma
2012-04-02 21:49 ` Tejun Heo
2012-04-02 22:03 ` Tao Ma
2012-04-02 22:17 ` Tejun Heo
2012-04-02 22:20 ` Tao Ma
2012-04-02 22:25 ` Vivek Goyal
2012-04-02 22:28 ` Tejun Heo
2012-04-02 22:41 ` Tao Ma
2012-04-03 15:37 ` IOPS based scheduler (Was: Re: [PATCH 18/21] blkcg: move blkio_group_conf->weight to cfq) Vivek Goyal
2012-04-03 16:36 ` Tao Ma
2012-04-03 16:50 ` Vivek Goyal
2012-04-03 17:26 ` Tao Ma
2012-04-04 12:35 ` Shaohua Li
2012-04-04 13:37 ` Vivek Goyal
2012-04-04 14:52 ` Shaohua Li
2012-04-04 15:10 ` Vivek Goyal
2012-04-04 16:06 ` Tao Ma [this message]
2012-04-04 16:45 ` Tao Ma
2012-04-04 16:50 ` Vivek Goyal
2012-04-04 17:17 ` Vivek Goyal
2012-04-04 17:18 ` Tao Ma
2012-04-04 17:27 ` Vivek Goyal
2012-04-04 18:22 ` Vivek Goyal
2012-04-04 18:36 ` Tao Ma
2012-04-04 13:31 ` Vivek Goyal
2012-03-28 22:51 ` [PATCH 19/21] blkcg: move blkio_group_conf->iops and ->bps to blk-throttle Tejun Heo
2012-03-28 22:51 ` [PATCH 20/21] blkcg: pass around pd->pdata instead of pd itself in prfill functions Tejun Heo
2012-03-28 22:51 ` [PATCH 21/21] blkcg: drop BLKCG_STAT_{PRIV|POL|OFF} macros Tejun Heo
2012-03-29 8:18 ` [PATCHSET] block: modularize blkcg config and stat file handling Jens Axboe
2012-04-02 20:02 ` Tejun Heo
2012-04-02 21:51 ` Jens Axboe
2012-04-02 22:33 ` Tejun Heo
2012-04-01 19:38 ` Vivek Goyal
2012-04-01 21:42 ` Tejun Heo
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=4F7C7176.5090504@tao.ma \
--to=tm@tao.ma \
--cc=axboe@kernel.dk \
--cc=cgroups@vger.kernel.org \
--cc=containers@lists.linux-foundation.org \
--cc=ctalbott@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=rni@google.com \
--cc=shli@kernel.org \
--cc=tj@kernel.org \
--cc=vgoyal@redhat.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;
as well as URLs for NNTP newsgroup(s).