LinuxQuestions.org - HDD high I/O rate?

- Linux - Newbie (https://www.linuxquestions.org/questions/linux-newbie-8/)

- - HDD high I/O rate? (https://www.linuxquestions.org/questions/linux-newbie-8/hdd-high-i-o-rate-4175498269/)

HDD high I/O rate?

Hello everyone!

It seems every time I begin to use my computer for more than an hour my HDD light begins to blink and it just won't turn off (meaning the HDD never stops spinning if i'm right). I decided to investigate this a little more and wanted to see what is using it currently. I decided to fire up iotop (a piece of software that shows you what is using the hard drive with a top-like interface) and this is what shows up at the very top:

Code:

 TID  PRIO  USER    DISK READ  DISK WRITE  SWAPIN    IO>    COMMAND 

 290 be/3 root        0.00 B/s    0.00 B/s  0.00 % 94.79 % [jbd2/sda1-8]

I honestly don't know what jbd2 is doing and why it is using 95% of the I/O. I find this kind of confusing expessially how it's not reading or writing anything to the disk. Side note that I built this computer myself so if anyone needs the specs of something I can normally call them off the top of my head :).

Again, I don't know if this is why the HDD light just won't quit blinking on my computer, but it's my only guess, and I did Google this problem and I did not get any relative results. If it helps, I am running Debian Wheezy (stable).

Thanks for anyone that helps!

From my quick googling, jdb2 is the journaling daemon for the ext4 filesystem and there are numerous reports for exactly the problem you describe(just google "jdb2" to see what I mean). The reported problems span a large time period and several different distors, see eg:

http://forums.debian.net/viewtopic.php?f=7&t=103450

https://bbs.archlinux.org/viewtopic.php?id=113516

http://forums.fedoraforum.org/showthread.php?t=290945

I don't see any definitive solution but I suspect it may be kernel related. You may want to review some of the suggestions in those threads.

....additionally whats missing here is the full kernel version, file system type, mount flags, `sysctl vm` values, a list of related configuration changes / customizations you made and output of `sysctl -w vm.block_dump=1; dmesg` (kill dmesg after a while, then zero the value).

Quote:

Originally Posted by kilgoretrout (Post 5135057)

Now that you bring up 'kernel related', I did notice that my kern.log seems to be quite large. Here is the output of 'ls -lS /var/log/*log | head':

Code:

-rw-r----- 1 root adm  1030372 Mar 15 18:33 /var/log/kern.log

-rw-r----- 1 root adm  868314 Mar 15 18:33 /var/log/daemon.log

-rw-rw-r-- 1 root utmp  292584 Mar 15 18:15 /var/log/lastlog

-rw-r----- 1 root adm  207835 Mar 15 18:33 /var/log/syslog

-rw-r--r-- 1 root root  200457 Mar 15 18:03 /var/log/pm-suspend.log

-rw-r--r-- 1 root root  188905 Mar 15 05:11 /var/log/dpkg.log

-rw-r--r-- 1 root root  183149 Mar 15 18:32 /var/log/Xorg.0.log

-rw-r--r-- 1 root root  124913 Mar 15 18:15 /var/log/pm-powersave.log

-rw-r----- 1 root adm    45221 Mar 15 18:22 /var/log/auth.log

-rw-r--r-- 1 root root  32064 Feb 24 02:50 /var/log/faillog

I took a look at the kern.log and it mostly sounds like it keeps printing stuff about my wireless card. Also, I was reading in the Fedora forum that someone had their install spanned over 2 hard drives. I have one hard drive with all the files on it.