
|
View Full Version : What is kjournald?
Alan - Vox 03-16-2002, 12:53 PM On all of my servers there abour 6 processes called kjournald, they are all running Red Hat Linux 7.2
On the other servers it just sits there idle, but on one of them it takes upto 20% of the cpu, anyone know what it is and why it takes so mcuh cpu time?
Softerweb 03-16-2002, 01:20 PM I would think that is the deamon for the journaling feature of EXT 3 filesystem. Check out http://www.redhat.com/support/wpapers/redhat/ext3/ for more info.
Daniel
mkaufman 03-16-2002, 03:23 PM Yes, it is for the journaling feature of the EXT3 file system. If it's taking up 20% of the resources (which it shouldn't..) a simple restart should probably fix it..
Alan - Vox 03-16-2002, 05:15 PM Ive already tried restarting the server, just came back.
131 root 17 0 0 0 0 SW 16.6 0.0 400:23 kjournald
23581 root 19 0 1080 1080 832 R 6.0 0.2 0:00 top
17843 root 12 0 588 548 488 S 4.9 0.1 38:39 syslogd
2082 named 11 0 34456 30M 15452 R 3.8 6.2 782:52 named
7 root 9 0 0 0 0 SW 2.2 0.0 203:44 kjournald
2080 named 11 0 34456 30M 15452 S 2.2 6.2 1627m named
AlaskanWolf 03-16-2002, 05:24 PM ext3 is known to take up alot of resources sometimes, specially on a bizy disk
suggest reverting back to ext2 if journaling is bothering u so much.
bitserve 03-17-2002, 02:45 AM I concur with what the alaskanwolf says. Are you using SCSI disks? Is there a lot of disk IO activity?
priyadi 03-17-2002, 06:51 AM If it is really caused by a lot of I/O activity, then I suggest to fine tune the journal for speed. See this url for more information about ext3: http://www.zip.com.au/~akpm/linux/ext3/.
If that doesn't help you, try converting to ext2, it should be as simple as updating your fstab file.
Alan - Vox 03-17-2002, 11:32 AM Yeah there is a most likely a lot of disk activity, its an ide disk.
Craig 03-17-2002, 11:43 AM Alan, in that quote of your processors, how long had the server been up?
Was that taken STRAIGHT after the restart or before?
Craig.
Alan - Vox 03-17-2002, 02:29 PM The server had been up a few days when i took that screen shot.
Originally posted by priyadi
If that doesn't help you, try converting to ext2, it should be as simple as updating your fstab file.
Sorry to post in an old thread.
But this daemon creates high loads on a few of my servers (an Ensim and a Cpanel). How someone can convert to ext2?
Is there much difference between ext2 and ext3 by means of safety and performance?
Regards,
Mac
priyadi 12-21-2002, 06:49 AM Originally posted by wmac
Sorry to post in an old thread.
But this daemon creates high loads on a few of my servers (an Ensim and a Cpanel). How someone can convert to ext2?
Is there much difference between ext2 and ext3 by means of safety and performance?
Regards,
Mac
No need to convert anything, you only need to edit /etc/fstab and change ext3 to ext2 then reboot the server (yes yes I know a reboot is not really necessary, but it is the easiest way). ext2 and ext3 are upward and backward compatible, the only difference is the journalled operation.
However I suggest you to try different approach first as ext2 requires lengthy fsck after a hard crash. It might take more than one hour for a moderately sized disk for today's standard.
High load on kjournald means you are having high data throughput. First try changing journalling mode to 'writeback', see man mount for more information. Second, try increasing journal size or maybe put the journal on another disk altogether, see man tune2fs for more information.
Brook 06-20-2008, 11:23 PM I'm getting high iowait and I have a 20 minute kjournald process running (and a 57 minute named process running :confused: )- should I change mine to writeback too?
Here's my fstab:
LABEL=/ / ext3 defaults,usrquota 1 1
LABEL=/boot /boot ext3 defaults 1 2
none /dev/pts devpts gid=5,mode=620 0 0
none /proc proc defaults 0 0
none /dev/shm tmpfs defaults 0 0
/dev/hda2 swap swap defaults 0 0
/usr/tmpDSK /tmp ext3 defaults,noauto 0 0
/tmp /var/tmp ext3 defaults,bind,noauto 0 0
If so which bits do I need to change?
Thanks in advance.
|