20030107: McIDAS: Data Update Problem (Again)

>From: Michael Keables <address@hidden>
>Organization: DU
>Keywords: 200301072011.h07KB0t25458 McIDAS

Hi Mike,

Happy New Year back atcha!

>The data files on cyclone.natnet.du.edu are not updating again (haven't
>since Dec 14) ...could you take a look at it when you get a second and see
>what is up?

No problem.

>Sorry to keep bothering you with this, but I am a one-man shop here at DU
>without a lot of unix system admin experience.

No worries.

OK, the problem is once again that /var is full:

/usr/local/ldm/decoders% df -k
Filesystem            kbytes    used   avail capacity  Mounted on
 ...
/dev/dsk/c0t0d0s1     413639  384656       0   100%    /var
 ...

Unfortunately, this time I have nothing under /var that I can delete
either as 'mcidas' or 'ldm'.  What you need to do is:

<as 'root'>
cd /var/tmp
<review the contents of the file 'contents'; if it is expendable, delete it.
If it needs to be kept, move it to a file system that has some space>

After making some room in /var, you will need to restart your LDM:

<as 'ldm'>
ldmadmin start

It is possible that the LDM queue has been corrupted, so it may be
advisable to start the LDM as follows:

ldmadmin stop
ldmadmin delqueue
ldmadmin mkqueue -f
ldmadmin start

I would have done all of the above myself, but I no longer have 'root'
access on cyclone.

Tom

>From address@hidden Wed Jan  8 13:42:08 2003
>Subject: RE: 20030107: McIDAS: Data Update Problem (Again)

Thanks, Tom.  Found a huge log file in /var/cron was the problem.

--MK