VSOM 7.6 - File system usage critcal but enough disk space avail

[Danny Meyer]

Hi.

On one of my combined VSOM/VSMS servers, I have a constant alert that says "File system usage on / critical".

But on the storage overview is enough free space, like 363G used / 1.2T total.

Any idea what could give me that error?

Thanks and regards

Danny

[Scott Olsen]

The storage overview only details the capacity and usage of the *Media Repositories* as far as I'm aware.

The actual root ("/") filesystem is considerably smaller and separate.  If you can, open an SSH session to the server (login 'localadmin') and check with:

df -h

If the usage on '/' is, in fact, significantly high, I'm willing to bet that:

/usr/BWHTTP/cores 

actually contains multiple dump files from core-msapi

Let me know!

[Danny Meyer]

Hi Scott.

Thanks very much for this good hint.

It really looks like sda2 is on its limit. But why and how to access it, do you have any hint?

con-physec-01:/ # df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/sda2              10G  9.1G  845M  92% /
udev                 1013M  156K 1013M   1% /dev
/dev/sda1              69M  7.4M   58M  12% /boot
/dev/sdb1             1.2T  353G  795G  31% /media1

The core directory is emtpy, so no dump files there.

Regards

Danny

[Scott Olsen]

Well... I'll tell you my quick and dirty method of tracking down space hogs via a shell;

du -BG --max-depth=1

... will list the disk usage in GB, for each directory, one level deep.  I'd start in /usr as it's likely log files or DB dumps related to BWHTTP ("VSM").  Find the largest subdirectory, change to it, re-run the du command.

If for some reason localadmin can't run it (it may not be in PATH), it's probably easier just to su to root for your hunting expedition:  sudo su -

Once you find out what it is, feel free to report back.  My second guess is 'bas' db dumps...

Cheers!

[Danny Meyer]

Hi Scott.

Your helpful linux commands are very good and I am sure many others will be glad to use as well.

It helped me to find some old files in /usr/src that were leftovers from the manual upgrade (its an msp unit). Deleted them and gained some extra space that made the alert go away. Not sure if the 10g partition will be helpful for the future, but I keep it since it is the standard size for the original suse setup (factory).

Thanks again for your always good advice.

Cheers

Danny

[gtoscano1]

Scott, 

thats slick, thanks for the tip.