V4R5M0 Upgrade to V5R2M0 uses 4GB

[Pete Helgren <Pete@xxxxxxxxxx>]

I upgraded a V4R5M0 Model 170 with 12GB of DASD over the weekend and was 
surprised to go from 54% to 89% disk utilization.  Can that be correct?

I also had a problem where I watched the % system ASP continue to climb 
until I did an ENDSBS *ALL at which point the % dropped back to about 
92%.  After applying the latest group PTF's I couldn't IPL due to lack 
of space (at least that is the error that displayed).  I re-IPL'd once 
again and managed to end all the subsystems before the ASP utilization 
reached 98% and then it dropped back to about 92% again.

I selectively started just a few subsystems/jobs and I am limping along 
OK now.  So here are my questions:

1.   Is the use of an additional 4GB of DASD normal for a V4R5M0 to 
V5R2M0 upgrade?
2.   If not, is there a specific place I should look for temporary 
objects I can jettison?
3.   How would you go about looking for the "runaway" job (at least I am 
assuming that the slow increase in %ASP use was due to a runaway job...).

It is a slightly old CUME (C4244520) (just ordered the latest) but I 
have a  model 800 at the same level and didn't experience the increasing 
% ASP utilization problem.

Ideas?

Pete Helgren