Re: Compile still running after 12+ hours?!?

On 08-Jul-2011 11:13 , Mike Wills wrote:

I walked away from a compile last night around midnight and today at
lunch (around noon) it was still being compiled when I killed it.
This is a 520 P5 system running 6.1. The RPG program uses SK's CSV
parser ( http://systeminetwork.com/article/how-read-csv-file-rpg ).
The only thing I know about with problems on the system is that the
admin is having problems getting the latest cume installed (working
with IBM on it) so we have the base 6.1 running right now. He thinks
he can get that installed finally this weekend. There was mentions in
the log of not finding some copybooks, but I have never seen that
hold up a compile, but I fixed those and tried again.

This isn't a one-time thing. It happens every time. The compile job
uses about 70% CPU so it seems to be doing something (can't imagine
what).

Has anyone seen this before? I'll be trying now after the cume is
installed to see if that is it.

Known defects for ILE RPG compiler looping problems should be included in a Google web search on the following:
APAR "loop" rpgle "cmpl" "6.1" site:ibm.com

One looping case is described by APAR SE45627 providing PTF R610 SI43988 which is not yet on a cumulative:
http://www-912.ibm.com/n_dir/nas4apar.NSF/c79815e083182fec862564c00079d117/a8be106177f11d20862577d0003c8a05?OpenDocument

Consider submitting the job HOLD(*YES), issuing STRSRVJOB on that job, then a minute or so after RLSJOB issue DMPJOBINT, then again after the job seems to level at the ~70% CPU utilization. A record of the LIC and compiler stack will be logged in each VLog produced by issuing that command against the serviced job. That information could be used to further narrow a search, or just left for comparison if the problem appears to remain after maintenance is eventually applied.

Regards, Chuck