Hi All,

I have /var mounted on system and it is occupied by 92% but unable to find what is used...

DF output:


root@segotl1712# df -Ph | grep /var
/dev/mapper/rootvg-varvol 5.8G 5.0G 501M 92% /var
/dev/mapper/postgres_vg01-soepostgresql 4.8G 12M 4.6G 1% /var/log/soepostgresql
/dev/mapper/postgres_vg01-pgsql_misc 9.8G 23M 9.2G 1% /var/lib/pgsql/misc
/dev/mapper/postgres_vg02-pgsql_xlog 9.8G 599M 8.7G 7% /var/lib/pgsql/xlog
/dev/mapper/postgres_vg03-pgsql_data 197G 5.6G 190G 3% /var/lib/pgsql/data
segote8813-07:/javad_isfsb_1000_q1001n 1.4T 695G 706G 50% /var/lib/pgsql/archive
root@segotl1712#


-------------------

DU ouput.

root@segotl1712# du -sh .* *
845G .
du: cannot access `../proc/23261/task/23261/fd/4': No such file or directory
du: cannot access `../proc/23261/task/23261/fdinfo/4': No such file or directory
du: cannot access `../proc/23261/fd/4': No such file or directory
du: cannot access `../proc/23261/fdinfo/4': No such file or directory
848G ..
4.0K account
257M cache
4.0K crash
4.0K cvs
1.4M db
8.0K empty
4.0K games
845G lib
4.0K local
24K lock
76M log
16K lost+found
0 mail
4.0K nis
4.0K opt
4.0K preserve
232K run
4.2M spool
144K tmp
4.0K yp
root@segotl1712#


==============================

Can anyone help on this issue.

Thanks in advance..


Regards,
Kathirvel.S

Well it's only 5.8G.
It's probably under your mountpoints.

Could you please elaborate and let me know how to fix this issue?.

Thanks.

Unmount these:Look in the mountpoints.

It is a production server.. Is there any other way to do without unmounting.

Check /lib, it says it has 845G !

Ok, just saw it /var that was full.
It seems you have /var on a separate lvm VolumeGroup: /dev/mapper/rootvg-varvol
The other ones seems like a different VG.
So don't grep for /var in the df command.

Probably (no clue about lvm) you can mount /dev/mapper/rootvg-varvol somewhere else (as well as where it is already mounted) and get access to the "base" filesystem.

Well, here rootvg is the Volume Group, to this one you add disks (or more correct partitions)
Inside it youe here have the Logical Volume varvol.
The space dedicated to this varvol is not same as the other Volume Groups.
postgres_vg0x onthe other hand is a different Volume Group using other disks or partitions, not same space.

So look in /var only for used space.
You can drill down like this:
du -s /var/*
du -s /var/dir-with-most-bytes/*

and so on

and do NOT unmount mounpoints for your postgres production database ! It have it's own disk and as you see plenty of room.

there can be inodes without filenames (for example still opened but already removed logfiles in /var/log). That will not be taken into account by du.

To look for large unlinked files:Scroll through the list and look for anything huge.

To look for files hidden under your active mount points:

1 members found this post helpful.

This worked well...Thanks.