linux-block.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Muneendra Kumar M <muneendra.kumar@broadcom.com>
To: linux-block@vger.kernel.org
Subject: number of write I/Os processed keeps changing for a same test case
Date: Tue, 5 Mar 2019 17:06:32 +0530	[thread overview]
Message-ID: <27226a7453efa3b147da2ea9f361437b@mail.gmail.com> (raw)

Hi ,

Iam running the below fio test case and collecting the cat
/sys/class/block/sda/stat before after the test.

And below are the max_sectors_kb  and max_hw_sectors_kb settings for the
device.

 cat /sys/class/block/sdb/queue/max_hw_sectors_kb

32767

[root@localhost ~]# cat /sys/class/block/sdb/queue/max_sectors_kb

1280

Test Case:
[root@localhost ~]#
io --filename=/dev/sdb --direct=1 --rw=write  --norandommap
--randrepeat=0 --ioengine=sync
 --bs=1m  --iodepth=1 --numjobs=1 --group_reporting --name=4ktest --size=50m

 And below is the o/p of the test .

[root@localhost ~]#cat /sys/class/block/sdb/stat

    6439        0  1513128     1618     4645        0  3332736     2697
0     3356      633        0        0        0        0

[root@localhost ~]#
io --filename=/dev/sdb --direct=1 --rw=write  --norandommap
--randrepeat=0 --ioengine=sync
 --bs=1m  --iodepth=1 --numjobs=1 --group_reporting --name=4ktest --size=50m

4ktest: (g=0): rw=write, bs=(R) 1024KiB-1024KiB, (W) 1024KiB-1024KiB, (T)
1024KiB-1024KiB, ioengine=sync, iodepth=1

fio-3.1

…..

Run status group 0 (all jobs):

  WRITE: bw=943MiB/s (989MB/s), 943MiB/s-943MiB/s (989MB/s-989MB/s),
io=50.0MiB (52.4MB), run=53-53msec



Disk stats (read/write):

  sdb: ios=0/0, merge=0/0, ticks=0/0, in_queue=0, util=0.00%

[root@localhost ~]# cat /sys/class/block/sdb/stat

    6519        0  1517272     1628     4795        0  3435136     2785
0     3420      633        0        0        0        0



And when I compare the difference between the number of write I/Os processed
between the two stats(4795-4645) it gave me 50 as it has submitted the 50
requests of 1MB each(since max_sectors_kb is set to 1280).

But when I ran the same test multiple the  number of write I/Os processed
(collected as above) keeps changing which is more than 50.

Not sure on what basis the requests are getting splitted when I run same
test case on the same setup.

Any input here will help me to understand more.

Iam running kernel  of 5.0.0-rc+ version.

Regards,
Muneendra

                 reply	other threads:[~2019-03-05 11:36 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=27226a7453efa3b147da2ea9f361437b@mail.gmail.com \
    --to=muneendra.kumar@broadcom.com \
    --cc=linux-block@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).