How to resolve filesystem full on CMS r12?

[soul0906]

I have experienced filesystem full (%iused = 100%) as below. I have tried to remove a lot of garbage but it still become full soon.

# df -F ufs -o i
Filesystem　　　　　 iused　　　 ifree　 %iused　 Mount
/dev/dsk/c0t0d0s0　 512000　　　　 0 　　100%　 /
/dev/dsk/c0t0d0s3　　 1887　　 382113　　 0%　 /cms

Other information as below:

# df -k
Filesystem 　　　　　 kbytes 　　 used　　 avail　　 capacity　Mounted on
/dev/dsk/c0t0d0s0 　 4130982　1260659　 2829014　　 31%　　 /
/proc 　　　　　　　　　　 0 　　　 　0 　　　 0 　　　0% 　　/proc
mnttab 　　　　　　　 　　0 　　　　 0 　　 　 0 　　　 0% 　　/etc/mnttab
swap 　　　　　　　 1112048 　　 　 24　 1112024 　　　1% 　　/var/run
/dev/dsk/c0t0d0s3　 3098743　　269774 　 2766995　　　 9%　　 /cms
swap 　　　　　　　 1112072 　 　 48　 1112024 　　　1% 　　/tmp

# uname -a
SunOS FT44430077 5.9 Generic_112233-11 sun4u sparc SUNW,Sun-Blade-100

# cat /etc/release
Solaris 9 4/03 s9s_u3wos_08 SPARC
Copyright 2003 Sun Microsystems, Inc. All Rights Reserved.
Use is subject to license terms.
Assembled 25 February 2003

[kamra]

How to resolve filesystem full on CMS r12?

soul0906,

Infact its the inodes in the '/' or root filesystem which are completely consumed i.e. 100% used. The inodes or index nodes are used for any new file creation at the OS level and once all are used, not even temporary files get generated and your CMS becomes at risk of getting hung until some of the files are deleted. This issue in CMS generally crops up due to incorrectly configured alarm subsystem and related dependencies in CMS, which cause generation of 0 byte files at a regular interval and this may go on for years before the inodes get full.
There may be other reasons as well for a large number of file creation consuming the inodes. This is better looked by T3/backbone team in Avaya, if you have no prior experience with Solaris/Unix OS and CMS application with such issues.

Saurabh Kamra