Re: Used Disk Capacity Jumped overnight to > 90%

On Tue, 1 Mar 2005 11:46:54 -0600, ron_adams@xxxxxxxxxxxxxx
<ron_adams@xxxxxxxxxxxxxx> wrote:
> We are on a new 520 at V5R3. Last night (Monthend) during processing,
> something happened on our system that bumped our used space from 80% to
> 97%!

Same thing happened on my system; the problem was a poorly designed
SQL query (did a  cartesian join of two huge files)  The QTEMP lib for
the job got huge fast.   Once the job ended, the system calmed down
and the usage went back to normal, but it took a few hours.  Are you
sure that all of the relevant jobs have completely ended?

What was different about this monthend?

> I have deleted numerous large objects/libraries, spool files, history,
> journals, etc. and I'm down to 91%, but I can't get down any further. IBM
> support has me running a RTVDSKINF to find out what's taking up all the
> space, but at the moment we're crippled. I suspect it has something to do
> with a Save/Restore of our Production databases last night, and the temp
> space not being released, but can't prove it until RTVDSKINF finishes.

So you have 11% that went up in smoke.  Can you march through the
monthend job stream and see what programs are doing what?   Perhaps a
program has changed since last month; it might be the culprit.

> Has anybody else had this or a similar experience? Any help that you can
> offer. I have deleted a lot of stuff already, but it doesn't seem to have
> helped.

You are obviously not deleting the right stuff ;->

I suspect you need a RCLSTG or something similar.  The RTVDSKING might
help, if you have a file with millions of deleted records.

-- 
Tom Jedrzejewicz
tomjedrz@xxxxxxxxx