[Mike Willeson]

Thanks Jim.  I agree with you.  Thanks Rob for the ideas.

I am not confident they will attempt to resolve their issue or change.  I will pursue handling the timeout on this side and move on to the next message.

As a side note, I will sympathize with the vendor a little bit as I found myself in a similar situation back in the late 90’s.  I was working for a vendor who was interfacing with a hospital where the “Clovertech board famous” Jim Cobane was working.

Their system was sending me a particular message our system just couldn’t digest and we were stuck.  I could not figure out how to move beyond it.  Jim removed the message and we were caught on a couple of subsequent messages for the same person which Jim also removed.  Problem solved.  

Until a couple months later when the exact same person became stuck again.  Jim removed the one message and we never had the issue surface again.

Jim may not remember it, but I remember it well.  So, Iguess whatever goes around comes around….    🙂

[Mike Willeson]

Well, I hear what you are saying, but am still unsure how the thread as a whole is to be configured.  I have a TCL under the client connection details that is correctly connecting and posting the details to the receiving system.

Is this TCL supposed to handle the ACK and if so, do I not configure the Inbound replies section of the Outbound tab?  I currently have that set to ‘Await Replies’ and have been trying to figure out what to use for an acknoweldgement handler.

Perhaps if you could offer an overview of how this thread is supposed to be configured or how the ACK Is supposed to work that might shed some light for me.

I apprecaite your help with all of my questions.

[Mike Willeson]

Thanks Arslan.

[Mike Willeson]

Worked perfectly!  Thanks Bob!

[Mike Willeson]

Tom,

I think Steve is pointing you in the right direction.  I have never done this with TCL before, but thinking about it from a database perspective, in your original values you were leaving the insert value for the prm_msgid value blank because you have an auto-incrementing primary key field there.

Now that you have added a value of ” to your insert statement, you are technically trying to insert a string in a numeric field.

It seems like you need to pre-determine an ID value you can use in your insert.  I am from the oracle world, and we had sequences we could use.  In SQL Server it appears there may be an “IDENTITY” you can use, but I don’t know how it works.

In psuedo code:

newID = select from IDENTITY to get numeric value

set insert “INSERT INTO PRM_PCIS_OrderNo_xRef (prm_msgid,pcis_orderno) VALUES(newID,’$msh_10′,”,”)”

Try a test and do like Steve suggested, have a variable for every insert parameter and make sure the variable is set to a real value of the proper data type.

[Author]

Replies