Scary (to me) messages from Co:Z

john.mckown · Post by **john.mckown** » Fri Jan 25, 2008 11:19 am

This is on z/OS 1.8 and Co:Z 1.0.4. I run the COZRUN batch job in the SAMPJCL. It runs successfully. However, on the z/OS SYSLOG, I get a number of "scary" messages. Most likely they are scary because I am not used to seeing them. These messages are not on the batch JES log or messages. I may be seeing them because my userid is RACF UAUDIT'ed.

<messages type="scary">
IEA989I SLIP TRAP ID=XEC6 MATCHED. JOBNAME=TSH009ZX, ASID=0038.
BPXP023I THREAD 0B5F545000000001, IN PROCESS 33554704, WAS 726
TERMINATED BY SIGNAL SIGINT, SENT FROM THREAD
0B5F454000000001, IN PROCESS 16777487, UID 1009, IN JOB TSH009ZX.
IEF450I TSH009ZX *OMVSEX - ABEND=SEC6 U0000 REASON=0000FF02 727
TIME=09.39.27
...
ICH408I USER(TSH009 ) GROUP(TSHG ) NAME(JOHN A MCKOWN ) 735
01000802 CL(IPCOBJ ) ID(0000065540)
INSUFFICIENT AUTHORITY TO W_GETIPC
ACCESS INTENT(R--) ACCESS ALLOWED(OTHER ---)
EFFECTIVE UID(0000001009) EFFECTIVE GID(0000001000)
</messages>

Post by **dovetail** » Fri Jan 25, 2008 11:53 am

John,

The SEC6 abend seems to be saying that one of the Co:Z processes was terminated using a SIGINT, which is true/normal but I'm not sure why it would result in an abend message. We don't see anything like this our our test systems.
I'll need to look into this to figure out why this might be displayed.

The RACF warning seems to be because you have auditing on. I'm not sure why there is a warning about "W_GETIPC" - we aren't using it but it might be used underneath the covers by the C library by the calls that we use to create processes and/or sockets. I'll try to do some research.

Can you run the launcher with tracing enabled (ARGS='LD,t ...') and email us the output along with the console messages? (please try to grab the syslog timestamp headers to help correlate). This might help to pinpoint which calls are causing which messages.

Thanks,
Kirk

Post by **dovetail** » Fri Jan 25, 2008 12:01 pm

Just a thought - w_getipc() might be getting called from the z/OS ssh client. Try running the just ssh client under z/OS and see if you get the RACF audit message.

To run the ssh client under z/OS, you probably need to use a shell under telnet or ssh (not OMVS) and do something like:

ssh id@host

john.mckown · Post by **john.mckown** » Fri Jan 25, 2008 4:09 pm

Good call! It is the ssh client which is doing the w_getipc and generating the messages. Now to see if I want to bother IBM with this. I don't think that the OpenSSH port is actually formally supported.

Post by **dovetail** » Fri Jan 25, 2008 5:03 pm

John,

ssh is included in the "Ported Tools for z/OS" program product, which is fully supported by the IBM support center.

If you are able to get any information about it why ssh is causing this warning message, please let us know.

Thanks again.
Kirk

Post by **dovetail** » Fri Jan 25, 2008 5:43 pm

John -

Are you still getting the SEC6 abend when during shutdown? I was hoping that this would stop when you got everything configured and running.

If not, please send us a trace (using -LD,t ) as described earlier and we will try to figure out which process it is and hopefully why you are seeing it.

Thanks again -

Kirk

john.mckown · Post by **john.mckown** » Thu Feb 07, 2008 9:59 am

Sorry about the delay in getting back. I was busy installing some new products. The problem has gone away now that I am properly configured. Everything is working well.