another BULKCPY problem

[Kevin Scantlan]

We recently upgraded our test system from 5.2 to 5.4.1  .  It took a little while to start having this problem.  I guess the right message to trigger this problem hadn’t been seen.  We had 2 orders interfaces aborting the inbound process.  I would leave the message in question in state 5.  Everytime I brought the process back up it would abort before brining any threads up until I removed the state 5 message.

I tested the translation with this same message and it would fail getting a segmentation error.  Support looked at it and suggested I remove the BULKCPY and insert appropriate PATHCPYs.  I did that and the error quit.  Both translations having problems were from a 2.3 variant to a 2.3 variant.  One, in fact, had the same variant for input and output.

The variant and translation were identical in both test and production (v 5.2).  I took the message in question, copied it to our production box and ran it with the test tool there.  The translations were still using the BULKCPY.  They had no problems there.

From this I would deduce that something was introduced with the upgrade code.

I hate to have to remove all my BULKCPYs from translations before I upgrade our production.  So far, it’s only happened on ORM messages, but how do I know it won’t happen elsewhere.  And it didn’t happen with all messages for those 2 interfaces, just some of them.

Anyone else have simiar problems.  I check old messages in this forum, but it seems they dealt with old variants.  There was a fix with rev 1 that was supposed to deal with 2.2 to 2.4 translations and BULKCPY.  But, it didn’t get a aborted process.

Topic