data=writeback & kjournald

A few questions on the same process. I have ext3 running on my server and it's taking up a whole lot of CPU at times due to the constant writing to the disk for email, MySQL database, & logs.

I searched for .journal and found there is a .journal file in each partition on my 2 drives.

/backup/.journal
/home/.journal
/usr/.journal
/var/.journal
/.journal

There are five occurances and each one hase a kjournald process runing constantly when I chaeck "top"

142 root 9 0 0 0 0 SW 3.3 0.0 32:48 kjournald
24340 siriusco 9 0 3076 2700 2144 S 0.7 0.2 0:00 proftpd
141 root 9 0 0 0 0 SW 0.5 0.0 28:27 kjournald
24299 host0711 11 0 1080 1080 820 R 0.5 0.1 0:00 top
32141 named 9 0 6428 4200 2448 S 0.1 0.4 0:23 named
24352 nobody 9 0 9556 8916 8760 S 0.1 0.9 0:00 httpd
24363 nobody 9 0 9556 8916 8760 S 0.1 0.9 0:00 httpd
24364 nobody 9 0 9572 8932 8776 S 0.1 0.9 0:00 httpd
1 root 7 0 508 464 440 S 0.0 0.0 1:20 init
2 root 9 0 0 0 0 SW 0.0 0.0 0:00 keventd
3 root 19 19 0 0 0 SWN 0.0 0.0 8:36 ksoftirqd_CPU0
4 root 9 0 0 0 0 SW 0.0 0.0 7:30 kswapd
5 root 9 0 0 0 0 SW 0.0 0.0 0:00 bdflush
6 root 9 0 0 0 0 SW 0.0 0.0 4:32 kupdated
9 root 9 0 0 0 0 SW 0.0 0.0 0:00 khubd
10 root 9 0 0 0 0 SW 0.0 0.0 25:26 kjournald
139 root 9 0 0 0 0 SW 0.0 0.0 4:34 kjournald
140 root 9 0 0 0 0 SW 0.0 0.0 27:39 kjournald
1154 root 9 0 588 544 488 S 0.0 0.0 0:37 syslogd
1159 root 9 0 1160 500 456 S 0.0 0.0 0:00 klogd
1215 nobody 9 0 2280 1568 1484 S 0.0 0.1 0:13 proftpd
1227 root 9 0 1092 984 912 S 0.0 0.1 0:08 sshd
1248 root 9 0 824 632 632 S 0.0 0.0 0:00 xinetd
1260 root 9 0 1352 1288 1164 S 0.0 0.1 6:40 antirelayd
1363 root 8 0 1096 928 868 S 0.0 0.1 0:07 exim
1377 root 8 0 684 656 616 S 0.0 0.0 0:10 crond
1387 root 9 0 1000 832 832 S 0.0 0.0 0:00 safe_mysqld
1442 mysql 9 0 12524 5584 1848 S 0.0 0.6 0:02 mysqld
1504 root 8 0 596 548 524 S 0.0 0.0 0:00 rhnsd
1537 root 9 0 760 740 644 S 0.0 0.0 0:00 portsentry
1541 root 9 0 760 740 644 S 0.0 0.0 0:00 portsentry
1543 mysql 9 0 12524 5584 1848 S 0.0 0.6 2:01 mysqld

These are always there and the CPU% for each will range from 0 to 3.0 and are always changing but together they add up to larg eloads many times a day for long periods sometimes.

My first question is does each partition need one of these? Or maybe just each disk? If I could cut it down to 2 it would save allot of CPU.

And my second question has to do with ext3 options. I read from http://www.redhat.com/support/wpaper...index.html#toc that I can set ext3 to "data=writeback" and i will cut down on the amount of logging and hopefully the CPU usage. But I checked my /etc/fstab file and there is no "data=" anything. I read that webpage and the man pages on mount and can't determine the correct process to get ext3 to use "data=writeback". Do I just need to add that exact text to the end of the /etc/fstab file using pico or do I have to do it with a command line using mount?

Thanks

 

 

 

 

Top