From: NeilBrown <neilb@suse.com>
To: "Jure Erznožnik" <jure.erznoznik@gmail.com>, linux-raid@vger.kernel.org
Subject: Re: Fwd: (user) Help needed: mdadm seems to constantly touch my disks
Date: Mon, 19 Dec 2016 09:21:08 +1100 [thread overview]
Message-ID: <87inqgvpyj.fsf@notabene.neil.brown.name> (raw)
In-Reply-To: <CAJ=9zifGPd+3YnP1aLwLPXjtpygRc9DsDP37CNz-VHBHn1vDOg@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 5506 bytes --]
On Thu, Dec 15 2016, Jure Erznožnik wrote:
> Thanks for helping Neil. I have run the suggested utilities and here
> are my findings:
>
> It is always [kworker/x:yy] (x:yy changes somewhat) or [0].
> A few lines from one of the outputs:
That's disappointing. "kworker" could be any work queue.
It might be useful to look for large scale patterns.
What different addresses are written to? Is there are regular pattern?
What is the period?
md doesn't use work_queues for IO so these much be coming from
elsewhere. bcache uses a few work-queues...
NeilBrown
>
> 9,0 3 0 0.061577998 0 m N raid5 rcw 3758609392 2 2 0
> 9,0 3 0 0.061580084 0 m N raid5 rcw 3758609400 2 2 0
> 9,0 3 0 0.061580084 0 m N raid5 rcw 3758609400 2 2 0
> 9,0 3 0 0.061580084 0 m N raid5 rcw 3758609400 2 2 0
> 9,0 3 0 0.061580084 0 m N raid5 rcw 3758609400 2 2 0
> 9,0 0 1 0.065333879 283 C W 11275825480 [0]
> 9,0 0 1 0.065333879 283 C W 11275825480 [0]
> 9,0 0 1 0.065333879 283 C W 11275825480 [0]
> 9,0 0 1 0.065333879 283 C W 11275825480 [0]
> 9,0 3 2 1.022155200 2861 Q W 11275826504 + 32 [kworker/3:38]
> 9,0 3 2 1.022155200 2861 Q W 11275826504 + 32 [kworker/3:38]
> 9,0 3 2 1.022155200 2861 Q W 11275826504 + 32 [kworker/3:38]
> 9,0 3 2 1.022155200 2861 Q W 11275826504 + 32 [kworker/3:38]
> 9,0 0 2 1.054590402 283 C W 11275826504 [0]
> 9,0 0 2 1.054590402 283 C W 11275826504 [0]
> 9,0 0 2 1.054590402 283 C W 11275826504 [0]
> 9,0 0 2 1.054590402 283 C W 11275826504 [0]
> 9,0 3 3 2.046065106 2861 Q W 11275861232 + 8 [kworker/3:38]
> 9,0 3 3 2.046065106 2861 Q W 11275861232 + 8 [kworker/3:38]
> 9,0 3 3 2.046065106 2861 Q W 11275861232 + 8 [kworker/3:38]
> 9,0 3 3 2.046065106 2861 Q W 11275861232 + 8 [kworker/3:38]
> 9,0 0 0 2.075247515 0 m N raid5 rcw 3758619888 2 0 1
> 9,0 0 0 2.075247515 0 m N raid5 rcw 3758619888 2 0 1
> 9,0 0 0 2.075247515 0 m N raid5 rcw 3758619888 2 0 1
> 9,0 0 0 2.075247515 0 m N raid5 rcw 3758619888 2 0 1
> 9,0 0 0 2.075250686 0 m N raid5 rcw 3758619888 2 2 0
> 9,0 0 0 2.075250686 0 m N raid5 rcw 3758619888 2 2 0
> 9,0 0 0 2.075250686 0 m N raid5 rcw 3758619888 2 2 0
> 9,0 0 0 2.075250686 0 m N raid5 rcw 3758619888 2 2 0
> 9,0 2 1 2.086924691 283 C W 11275861232 [0]
> 9,0 2 1 2.086924691 283 C W 11275861232 [0]
> 9,0 2 1 2.086924691 283 C W 11275861232 [0]
> 9,0 2 1 2.086924691 283 C W 11275861232 [0]
> 9,0 0 3 2.967340614 1061 Q FWS [kworker/0:18]
> 9,0 0 3 2.967340614 1061 Q FWS [kworker/0:18]
> 9,0 0 3 2.967340614 1061 Q FWS [kworker/0:18]
> 9,0 0 3 2.967340614 1061 Q FWS [kworker/0:18]
> 9,0 3 4 3.070092310 2861 Q W 11275861272 + 8 [kworker/3:38]
> 9,0 3 4 3.070092310 2861 Q W 11275861272 + 8 [kworker/3:38]
> 9,0 3 4 3.070092310 2861 Q W 11275861272 + 8 [kworker/3:38]
> 9,0 3 4 3.070092310 2861 Q W 11275861272 + 8 [kworker/3:38]
> 9,0 0 0 3.101966398 0 m N raid5 rcw 3758619928 2 0 1
> 9,0 0 0 3.101966398 0 m N raid5 rcw 3758619928 2 0 1
> 9,0 0 0 3.101966398 0 m N raid5 rcw 3758619928 2 0 1
> 9,0 0 0 3.101966398 0 m N raid5 rcw 3758619928 2 0 1
> 9,0 0 0 3.101969169 0 m N raid5 rcw 3758619928 2 2 0
> 9,0 0 0 3.101969169 0 m N raid5 rcw 3758619928 2 2 0
> 9,0 0 0 3.101969169 0 m N raid5 rcw 3758619928 2 2 0
> 9,0 0 0 3.101969169 0 m N raid5 rcw 3758619928 2 2 0
> 9,0 0 4 3.102340646 283 C W 11275861272 [0]
> 9,0 0 4 3.102340646 283 C W 11275861272 [0]
> 9,0 0 4 3.102340646 283 C W 11275861272 [0]
> 9,0 0 4 3.102340646 283 C W 11275861272 [0]
> 9,0 3 5 4.094666938 2861 Q W 11276014160 + 336
> [kworker/3:38]
> 9,0 3 5 4.094666938 2861 Q W 11276014160 + 336
> [kworker/3:38]
> 9,0 3 5 4.094666938 2861 Q W 11276014160 + 336
> [kworker/3:38]
> 9,0 3 5 4.094666938 2861 Q W 11276014160 + 336
> [kworker/3:38]
> 9,0 3 0 4.137869804 0 m N raid5 rcw 3758671440 2 0 1
> 9,0 3 0 4.137869804 0 m N raid5 rcw 3758671440 2 0 1
> 9,0 3 0 4.137869804 0 m N raid5 rcw 3758671440 2 0 1
> 9,0 3 0 4.137869804 0 m N raid5 rcw 3758671440 2 0 1
> 9,0 3 0 4.137872647 0 m N raid5 rcw 3758671448 2 0 1
> 9,0 3 0 4.137872647 0 m N raid5 rcw 3758671448 2 0 1
> 9,0 3 0 4.137872647 0 m N raid5 rcw 3758671448 2 0 1
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]
next prev parent reply other threads:[~2016-12-18 22:21 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-12-13 8:13 (user) Help needed: mdadm seems to constantly touch my disks Jure Erznožnik
2016-12-14 1:15 ` NeilBrown
[not found] ` <CAJ=9zieRuTNiEGuB_RouqbdLGoxNkn09yiogR6rND84LtMdbxA@mail.gmail.com>
2016-12-15 7:01 ` Fwd: " Jure Erznožnik
2016-12-18 19:40 ` Jure Erznožnik
2016-12-18 21:30 ` Theophanis Kontogiannis
2016-12-18 22:21 ` NeilBrown [this message]
[not found] ` <CAJ=9zidNV4sPj7KC7_mJEo8+=-YTKyWD5RiLsGG9p33CV12Qdg@mail.gmail.com>
2016-12-19 4:01 ` Fwd: " NeilBrown
2016-12-19 7:12 ` Jure Erznožnik
2016-12-19 23:39 ` NeilBrown
2016-12-21 11:33 ` Jure Erznožnik
2016-12-21 22:52 ` NeilBrown
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=87inqgvpyj.fsf@notabene.neil.brown.name \
--to=neilb@suse.com \
--cc=jure.erznoznik@gmail.com \
--cc=linux-raid@vger.kernel.org \
/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).