Flexible I/O Tester development
 help / color / mirror / Atom feed
From: Erwan Velu <erwan@seanodes.com>
To: fio@vger.kernel.org
Subject: Help required to understand fio output
Date: Wed, 19 Nov 2008 14:15:10 +0100	[thread overview]
Message-ID: <4924115E.6040701@seanodes.com> (raw)

Here come a typical output I have using fio. I added some line numbers 
as an helper.

01: /dev/mydisk: (groupid=1, jobs=1): err= 0: pid=7433
02:  write: io=79,112MiB, bw=72,774KiB/s, iops=740, runt=1139888msec
03:    slat (usec): min=9, max=572, avg=21.22, stdev= 7.03
04:    clat (msec): min=2, max=668, avg=86.43, stdev=24.47
05:    bw (KiB/s) : min=    0, max=80749, per=100.06%, avg=72815.92, 
stdev=2318.78
06:  cpu          : usr=0.93%, sys=1.35%, ctx=722885, majf=0, minf=10031
07:  IO depths    : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=0.1%, 32=0.1%, 
 >=64=100.0%
08:     submit    : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, 
 >=64=0.0%
09:     complete  : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.1%, 
 >=64=0.0%
10:     issued r/w: total=0/843855, short=0/0
11:
12:     lat (msec): 4=0.01%, 10=0.03%, 20=0.38%, 50=6.53%, 100=64.24%
13:     lat (msec): 250=28.79%, 500=0.01%, 750=0.01%


I have some questions about the output.

About Latencies:
Line 4, clat is reported to be at 86ms as average but line 12 is is 
written that 100=64 250=29
What's the meaning of "100=" ? Does it means "there were 64% of IOs 
completed between 100 & 249 ms ?"
If yes,  that's pretty surprising that while having 95% of my IOs that 
are completed in more than 100ms, the average is reported as 86.
If "100=" means "IO between 50 and 99ms", thus output become pretty 
confusing.

About IO depths:
In my config file I had set a depth to 64. Line 07 reports that 100% of 
my IOs where >=64... cool but is it also said that 0.6% were spent on <=64K
Sounds to be a rounding effect but does this really means that some IOs 
where sent with a short depth ?
And then, how should I understan lines 8 & 9 that reports that 100% of 
IOs were completed with a depth=4... What happened ?

Hope this questions aren't too stupid ;o)
Thanks,

-- 
Erwan Velu
Pre-Sales Engineer
Seanodes
http://www.seanodes.com
+33 (0)1 41 22 13 83


             reply	other threads:[~2008-11-19 13:15 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-11-19 13:15 Erwan Velu [this message]
2008-11-19 13:19 ` Help required to understand fio output Jens Axboe
2008-11-19 13:42   ` Erwan Velu
2008-11-19 13:43     ` Jens Axboe

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=4924115E.6040701@seanodes.com \
    --to=erwan@seanodes.com \
    --cc=fio@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