Editing Routine Maintenance (section)

== Free up space on Other Servers ==

Many servers start to run out of disk space over time.  Often it is caused by unread mail for
root or log files.  

To find the source of the problems, you use "du" to find where the disk space is being used.
You can't do a du on /proc or /dev, so I use the command

<pre>
[root@virt11 /]# du -hs [a-c]* deprecated [e-o]* [q-u]* var | tee duhs0

which produces something like this.

4.0K    backup
4.0K    backup1
4.0K    backup2
4.0K    backup3
4.0K    backup4
7.5M    bin
47M     boot
4.0K    deprecated
92M     etc
30M     home
8.0K    initrd
541M    lib
16K     lost+found
8.0K    media
0       misc
8.0K    mnt
0       net
92M     opt
336K    root
36M     sbin
8.0K    selinux
8.0K    srv
0       sys
4.0K    test
16K     tmp
1.2G    usr
583M    var

In this case it looks like /var is the problem, so

cd /var
du -hs * | tee duhs9

Produces

12K     account
2.6M    analog-5.32
63M     cache
24K     db
4.0K    duhs
4.0K    duhs1
4.0K    duhs2
4.0K    duhs3
4.0K    duhs4
4.0K    duhs5
4.0K    duhs6
4.0K    duhs7
4.0K    duhs8
32K     empty
8.0K    games
16K     kerberos
42M     lib
8.0K    local
36K     lock
457M    log
0       mail
8.0K    nis
8.0K    opt
8.0K    preserve
8.0K    racoon
240K    run
18M     spool
8.0K    tmp
64K     vz
0       vzagent
0       vzagent.tmp
16K     vzquota
1.2M    www
20K     yp

Usually, the problem is in /var/spool or /var/log, due to unread mail
or excessive log files.  You can continue to drill down by doing
a "cd <sbdirectory>" and another "du -hs *".

</pre>