Interpreting /proc/diskstats

rajesht · 09-05-2005, 07:30 AM

HI,
I need to interpret the entries in the file "/proc/diskstats". I searched and found about the first three fields. The first three fields are "Major Device Number", "minor Device Number" and "Device Name". Could anyone help me interpret the rest of the fields.

Thanks.

Rajesh.

Noth · 09-06-2005, 11:48 PM

Take a look at Documentation/iostats.txt in a kernel source tree.

rajesht · 09-08-2005, 07:21 AM

I have searched throughout my box and was not able to find the file "iostat.txt". could you please help me.

Noth · 09-08-2005, 10:27 AM

It's iostats.txt, plural. And it's in the kernel source tree, if you haven't installed any kernel-source packages or downloaded the source from kernel.org you probably won't have one.

cbellamk · 07-15-2009, 02:03 AM

Field 1 -- # of reads issued
Field 2 -- # of reads merged, field 6 -- # of writes merged
Field 3 -- # of sectors read
Field 4 -- # of milliseconds spent reading
Field 5 -- # of writes completed
Field 7 -- # of sectors written
Field 8 -- # of milliseconds spent writing
Field 9 -- # of I/Os currently in progress
Field 10 -- # of milliseconds spent doing I/Os
Field 11 -- weighted # of milliseconds spent doing I/Os

#cat /proc/diskstats | grep hda

3 0 hda 43205 4113 4800428 280967 1051597 1682874 21876608 1950120 0 858685 2231096
3 1 hda1 25838 525266 1505217 12041736
3 2 hda2 846 1164 88 704
3 3 hda3 20498 4272886 1229274 9834168

Let`s consider "How I can get stats for my /dev/hda3 ??"
Let`s consider this line : 3 3 hda3 20498 4272886 1229274 9834168
After a word "hda3" we have four 32-bit (type integer) fields, us interesting these:
The second field: 4272886 (total number of sectors which normally reads from /dev/hda3)
and The Fourth filed : 9834168 (total number of sectors which normally writes to /dev/hda3)
These two fields - all that is necessary for us. A /proc/diskstats continuously updated and all that is necessary for us - make measurements for "second field" and "fourth field" in two different moment of time, receiving a difference of values and dividing it into an interval of time, we shall have Disk I/O stats in sectors/sec. Multiply this result on 512 (number of bytes in one sector) we shall have Disk I/O stats in bytes/sec.
May be exist a problem how to measure precisely moment of a time in seconds ...
Here a piece of my code on С for this purpose:

gacanepa · 12-30-2013, 01:34 PM

First off, I apologize for replying to an old thread, but I thought that someone might find this useful:

Quote:

What: /proc/diskstats
Date: February 2008
Contact: Jerome Marchand <jmarchan@redhat.com>
Description:
The /proc/diskstats file displays the I/O statistics
of block devices. Each line contains the following 14
fields:
1 - major number
2 - minor mumber
3 - device name
4 - reads completed successfully
5 - reads merged
6 - sectors read
7 - time spent reading (ms)
8 - writes completed
9 - writes merged
10 - sectors written
11 - time spent writing (ms)
12 - I/Os currently in progress
13 - time spent doing I/Os (ms)
14 - weighted time spent doing I/Os (ms)
For more details refer to Documentation/iostats.txt

Source: https://www.kernel.org/doc/Documenta...ocfs-diskstats

larchicha · 02-23-2014, 05:58 AM

Thank you gacanepa !
You've been of great help... you're the first one i found explaining the true meaning of all these numbers... (after a good 15min google search

)

I wish there was a like button, you made my day.

syg00 · 02-23-2014, 06:37 AM

There is - 2 actually.
Click the "Yes" after "Did you find this post useful?", or if you want to leave a comment, click the "rep" link under the userid on the left.

Oh, and welcome ...

larchicha · 02-25-2014, 03:10 PM

Didn't see that... Thank you.

And Thanks again for the welcome, i always end myself on this website after searching,
without ever posting or thanking... what a rude person i am