How to measure IOPS

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


 > Can you give me a hint what fio parameters you would suggest to measure IOPS?

I would put your workload in a parameterized file and use the iodepth
as a parameter (or other ones like bs).

For example like this:

iodepth=${IODEPTH} (in the job file)

Then a script:

for i in `seq 1 10` # Repeat runs to observe variation.
 for depth in 1 2 4 8 16 32 # Increase iodepth
  IODEPTH=$depth fio <job file>

I also would use options like write_lat_log to visualize what is

The goal would be to obtain, not a single iops number, but maximum,
minimum, average and their variation depending on the chosen
parameter and see what seems to be a stable result.

There are probably other useful things to do (I am new to fio, too).

 > Are the 140 IOPS too high for a 7.200k drive like the one I have?

I would have a look at reviews of similar disks for similar tests.
For example performance database for
IOMeter tests.
There is an example fio job (iometer-file-access-server) which
mimics this test.

Some remarks:

 > [root@iotest ~]# dd if=/dev/zero of=/dev/sdd bs=512 count=1000 oflag=direct

I have observed variability in dd tests, you may want to do such tests
in a loop.  There are several others dd in sg3_utils, you may want to
try some and see if you get similar results.

 >   sdd: ios=0/1467, merge=0/1, ticks=0/651445, in_queue=651760, util=99.16%

I would use a test which drives util to 100%

To unsubscribe from this list: send the line "unsubscribe fio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at

[Home]     [Linux SCSI]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Photo]     [Yosemite News]    [Yosemite Photos]    [Video Projectors]     [Free Online Dating]     [Linux Kernel]     [Linux SCSI]     [XFree86]

Powered by Linux