Page 1 of 1

strange CICS programs abend

PostPosted: Thu Mar 17, 2011 6:05 pm
by aqswde
Hello,

We a strange sequence of operations which cause to abend.
The sequence is like this :
call to program X -> Call to program Y -> Call to program Z -> PF3 to program Y -> PF3 to program X -> ABEND
sequence that works :
call to program X -> Call to program Y -> PF3 to program Y -> PF3 to program X
and
call to program X -> Call to program Y -> call to program T(diff from Z) -> PF3 to program T-> PF3 to program Y -> PF3 to program X

we check to commarea and I'ts look OK, but this specific senerio cause abend inspite that wee see that program Y save program X to get back to in PF3.

The abend not points about something specific, the abend:
" Program Y, EZERESRC-CLEAN a CICS READ error occured in service ELAESERR
CICS EIBFN 0602, RCODE 0C0000000000, RESP 19, RESP2 60
an error occurred in Program Y on statement number 348
an error occurred starting transaction Program X
CICS EIBFN 5002, RCODE 0000001C0000, RESP 28, RESP 21"

Thanks in advance!

Re: strange CICS programs abend

PostPosted: Thu Mar 17, 2011 6:26 pm
by Robert Sample
As you posted your question, it cannot be answered.

When you write "call to program <?>", do you mean
- EXEC CICS LINK
- EXEC CICS XCTL
- program language CALL (which language? static or dynamic call?)

These are not handled the same way by the system, so without knowing at least this basic information your question has no answer.

Re: strange CICS programs abend

PostPosted: Thu Mar 17, 2011 6:43 pm
by BillyBoyo
In addition to Robert, does the "statement number 348" give you any clues?

I assume the error is consistent, ie always repeatable. Looks like Program Z is upsetting something. If you just go in-and-out, to minimise the amount of code to look at, is there something Z does that T doesn't, or vice versa? Do you get a dump? If so, does the program code in Y still look like program code? What does all the information in the error messages mean to you? Any clues there? At the moment, you have the easy part. We can spend all day guessing...

Re: strange CICS programs abend

PostPosted: Fri Mar 18, 2011 8:16 pm
by stevexff
The way you describe it, this is a conversational program rather than a psuedoconversational program and nobody except for IBM ("do as I say, not as I do") ever writes those in CICS because they hold resources between interactions.

Re: strange CICS programs abend

PostPosted: Fri Mar 18, 2011 8:43 pm
by mongan
Have you followed your CICS flow with CEDF?