Accuracy of disk statistics IO counter
Awhile back I had suggested moving the place in the block driver logic where stats get updated to more accurately reflect what was happening at the time they were actually sent to the drivers and I believe they were indeed made in the 2.6.15 timeframe. I've recently started taking closer look at the numbers and while the sector counts look correct I'm not sure I'd agree with the number of I/Os being reported and believe they can be off by maybe 15% or more.
Specifically, I wrote a 1GB file with a blocksize of 1MB, which would result in 1000 writes at the application level. What I believe then happens is that each write turns into 8 128KB requests to the driver, which should result in 8000 actual writes. Toss in metadata operations and who knows what else and the actual number should be a little higher. What I've see after repeating the tests a number of times on 2.6.16-27 is numbers ranging from 6800-7000 writes which feels like a big enough difference to at least point out.
-mark
Quoted 8 lines · Mark Seger
Awhile back I had suggested moving the place in the block driver logic where stats get updated to more accurately reflect what was happening at the time they were actually sent to the drivers and I believe they were indeed made in the 2.6.15 timeframe. I've recently started taking closer look at the numbers and while the sector counts look correct I'm not sure I'd agree with the number of I/Os being reported and believe they can be off by maybe 15% or more.
Quoted 8 lines
Specifically, I wrote a 1GB file with a blocksize of 1MB, which would result in 1000 writes at the application level. What I believe then happens is that each write turns into 8 128KB requests to the driver, which should result in 8000 actual writes. Toss in metadata operations and who knows what else and the actual number should be a little higher. What I've see after repeating the tests a number of times on 2.6.16-27 is numbers ranging from 6800-7000 writes which feels like a big enough difference to at least point out.
Install http://brick.kernel.dk/snaps/blktrace-git-20060723022503.tar.gz
and blktrace your disk for the duration of the test and compare the io numbers. Requires 2.6.17 or later, though.
Signature · 1 line
Jens Axboe
Quoted 10 lines
Quoted 4 lines
Specifically, I wrote a 1GB file with a blocksize of 1MB, which would result in 1000 writes at the application level. What I believe then happens is that each write turns into 8 128KB requests to the driver, which should result in 8000 actual writes. Toss in metadata operations and who knows what else and the actual number should be a little higher. What I've see after repeating the tests a number of times on 2.6.16-27 is numbers ranging from 6800-7000 writes which feels like a big enough difference to at least point out.
Install http://brick.kernel.dk/snaps/blktrace-git-20060723022503.tar.gz
and blktrace your disk for the duration of the test and compare the io numbers. Requires 2.6.17 or later, though.
I had problems getting blktrace going and posted a note to [email protected] at Alan Brunelle's suggestion and he also said he'd take a closer look at it himself. On the other hand he pointed me to 'stap' and I was able to use it to get the details I was looking for - SystemTAP really rocks for this type of analysis! As it turns out my assumption about driver blocksize was wrong (sorry about that). It turns out that the size of requests from the driver to the disk is 160mb and so the I/O count was smaller than I had anticipated. -mark
Quoted 12 lines · Mark Seger
Quoted 12 lines
Quoted 11 lines
Specifically, I wrote a 1GB file with a blocksize of 1MB, which would result in 1000 writes at the application level. What I believe then happens is that each write turns into 8 128KB requests to the driver, which should result in 8000 actual writes. Toss in metadata operations and who knows what else and the actual number should be a little higher. What I've see after repeating the tests a number of times on 2.6.16-27 is numbers ranging from 6800-7000 writes which feels like a big enough difference to at least point out.
Quoted 7 lines
Quoted 5 lines
Install http://brick.kernel.dk/snaps/blktrace-git-20060723022503.tar.gz
and blktrace your disk for the duration of the test and compare the io numbers. Requires 2.6.17 or later, though.I had problems getting blktrace going and posted a note to [email protected] at Alan Brunelle's suggestion and he also
Documentation issues, sorry about that. It hasn't been updated for the debugfs change.
Quoted 6 lines
said he'd take a closer look at it himself. On the other hand he pointed me to 'stap' and I was able to use it to get the details I was looking for - SystemTAP really rocks for this type of analysis! As it turns out my assumption about driver blocksize was wrong (sorry about that). It turns out that the size of requests from the driver to the disk is 160mb and so the I/O count was smaller than I had anticipated.
160KiB, I'm assuming? :-)
Signature · 1 line
Jens Axboe
Quoted 4 lines
160KiB, I'm assuming? :-)
oops... 8-(
-mark