Void Journal BSFN not writing a reversing F0911

[jimmymac]

I support two OneWorld XE environments, and have begun haveing a specific problem related to Voiding a Service Billing Invoice. One of these environments was recently created and since then this problem has occurred only in the new environment. We are running on two ISeries V5R4, OneWorld XE Sp23.

When voiding a service billing invoice, several updates are supposed to be performed. Voiding the F03B11, Voiding the entries in the F4822, and if the GL entries have been posted, F0911 records should be written to reverse the orginal F0911 entries. In our new environment only, when a user voids the invoice, all of the updates are performed, except in most instances, the F0911 to reverse the transaction is not written, causing the batch to be out of balance. The user has to manually enter the transaction to post the batch.

Debugging the issue, the problem appears to be in the B0900026, the function that reads the original F0911 and inserts the new F0911. I have debugged this and it fails due to a fetch of the F0911 not retreiving the original records. So the While loop that will do the insert isn't entered. I have debugged this when its worked and when it has not, and when it has not, the failed fetch didn't make any sense. The records are there, the fetch just fails and the while doesn't execute.

In our two environments, this is working perfectly in our old environment, and is failing in our new environment about 80% of the time. I had a thought to rebuild the indexes over the F0911 thinking that an index is somehow corrupt.

If this were a batch job, I would delete the SQLPKG, but this isn't.

Any thoughts as to where to look would be appreciated. I have turned on debugging and do not see any difference between a log where it works and one where it doesn't, except the fail of the fetch that really should work.

Thanks.

 

[llallen1113]

Have you verified that your bsfn OCM mappings are the same? Sometimes, business functions really need to be set to run locally, even though it's not specified as such on the business function location field in bsfn design. Also, check if you have a default bsfn mapping, and verify that the default is the same.

 

[ramsib1]

I had faced similar issue with advanced pricing when using V5R3 on iSeries. In the end it turned out to be IBM issue. We called them over and they app lied some PTFs and things got sorted out.






To: [email protected]
From: [email protected]
Subject: Void Journal BSFN not writing a reversing F0911
Date: Thu=2C 22 Oct 2009 10:31:19 -0500

I support two OneWorld XE environments=2C and have begun haveing a specific problem related to Voiding a Service Billing Invoice. One of these environ ments was recently created and since then this problem has occurred only in the new environment. We are running on two ISeries V5R4=2C OneWorld XE Sp2 3.

When voiding a service billing invoice=2C several updates are supposed to b e performed. Voiding the F03B11=2C Voiding the entries in the F4822=2C and if the GL entries have been posted=2C F0911 records should be written to re verse the orginal F0911 entries. In our new environment only=2C when a user voids the invoice=2C all of the updates are performed=2C except in most in stances=2C the F0911 to reverse the transaction is not written=2C causing t he batch to be out of balance. The user has to manually enter the transacti on to post the batch.

Debugging the issue=2C the problem appears to be in the B0900026=2C the fun ction that reads the original F0911 and inserts the new F0911. I have debug ged this and it fails due to a fetch of the F0911 not retreiving the origin al records. So the While loop that will do the insert isn't entered. I have debugged this when its worked and when it has not=2C and when it has not =2C the failed fetch didn't make any sense. The records are there=2C the fe tch just fails and the while doesn't execute.

In our two environments=2C this is working perfectly in our old environment =2C and is failing in our new environment about 80% of the time. I had a th ought to rebuild the indexes over the F0911 thinking that an index is someh ow corrupt.

If this were a batch job=2C I would delete the SQLPKG=2C but this isn't.

Any thoughts as to where to look would be appreciated. I have turned on deb ugging and do not see any difference between a log where it works and one w here it doesn't=2C except the fail of the fetch that really should work.

Thanks.