MVSFORUMS.com

saptarshi_tcs · Beginner Joined: 29 Jul 2005 Posts: 3 Topics: 1

I am facing an AFCY Abned while writng in a KSDS VSAM file.
The file has no alternate Indexes.
The situation is like
1. prog1 creates a TSQ (which can have atmost 1222 items). It calls Prog2 after builing the TSQ.
2. Prog2 reads each items in the TSQ sequntially, and for each read it calls prog3 with the TSQ data.
3. Prog3 with that data Read-Updates a VSAM file the it reads another VSAM file with that data and if the read is successful then it writes to another VSAM file. In this last write there is a AFCY abend.

It is occuring in production quite often and the problem is in Test region we are unable to simulate the abend. With expeditor it is working fine.

From the Joblog I found the message-
'Threshold TRANS CPUTIME FHPL 0153 PROBLEM V= 10.838 W= 8.500 P= 9.000 '

I require suggestions on how to crack this problem.

kolusu · Posted: Fri Jul 29, 2005 5:50 am Post subject:

saptarshi_tcs,

check this for a detailed explanation of the abend

saptarshi_tcs · Beginner Joined: 29 Jul 2005 Posts: 3 Topics: 1

Thanks for the reply.
The explanation is fine; but how to resolve the issue.
As we can't simulate the abend in Test and from expeditor it seems the program is working fine means it is not getting into any sort of loop.
So, my point is because there are so many write request say around 2500 records are wriiten at the particular instance when the abend occured, can there be a possibility that the Free Space definition should be changed in the file description because there are random writes in the file.

Please provide suggestions on how to resolve this issue to avoid any such abends in future.

Manas Biswal · Posted: Mon Aug 01, 2005 10:06 am Post subject:

saptarshi_tcs,

Please contact your CICS Sys admin. Although your message does not tell much, but it seems that your transaction is getting timed out and purged, while it is still writing or waiting to write to the VSAM file. There are values that you can specify for the DTIMEOUT parameter, which control how much time a transaction can remain suspended before getting timed out. But it seems from your mail, that your transaction is active and not in a suspended state. So, I am not sure of whether this parameter is the cause of your problem.
What I can certainly say that this is an issue for your CICS Sysadmin.

For now, can you please provide us with the relevant messages from the CICS IPL job of your CICS region.

Thanks,
Manas
_________________
There is no path to peace. Peace is the path.
- Mahatma Gandhi (1869-1948)

saptarshi_tcs · Beginner Joined: 29 Jul 2005 Posts: 3 Topics: 1

Hi,

What I found from the CICS job log is
'Threshold TRANS CPUTIME FHPL 0153 PROBLEM V= 10.838 W= 8.500 P= 9.000 '

where W is the time after which a warning is given and
P is the time after which CICS issues an Purge and
V the actual time taken by the Transaction.

Anyways, I am guessing that it has become a performance issue with the VSAM file on which the write operation was requested.

For that VSAM file(after the abend day before del/def) I found out that there were some 6000 CI splits and 36 CA splits. 94% of its extents were used.

Can this be the reason for such kind of abends.

Manas Biswal · Posted: Mon Aug 01, 2005 12:38 pm Post subject:

Is there nothing else on the CICS job log. Some DFH message or anything like that. Pls post the complete message list pertaining to the transaction.

Thanks,
Manas
_________________
There is no path to peace. Peace is the path.
- Mahatma Gandhi (1869-1948)