MVSFORUMS.com

[THRIVIKRAM]

Hi All,

We have MQ triggering the application program-1 when the first message comes on to the MQ. This program-1 calls program-2. We are using external copybooks to pass the MQ data from program-1 to program-2. The MQ data can be upto 4Meg in length. What is happening here is, not all the records will be 4Meg but we are reserving 4M space. CICS folks told us that we need to do some fix to minimize the getmains.
My idea is to define external Copybook like this:
01 External-Header
05 Input-MQ-Msg Occurs 1 to 4M depending on A-record-length
10 Input-byte-record Pic X(1)

While performing MQ-Get,
Move 4M to BufferLength.
MQGET (Hconn, Hobj, MsgDesc, GetMsgOpts, BufferLength,
Buffer, A-record-length, CompCode, Reason)

Does this command use only required amount of bytes in external copybook..?

Thanks!!

[dbzTHEdinosauer]

BufferLength defines the buffer size.
_________________
Dick Brenholtz
American living in Varel, Germany

[RonB]

We had a "similar" requirement and solved it by opening the MQ twice. In the first open, we open the MQ in browse mode and MQGET the message into a short ( 2K ) data area ( just large enough to hold the XML-Header information plus some application header info ) with the MQGMO_BROWSE_FIRST/MQGMO_BROWSE_NEXT and MQGMO_ACCEPT_TRUNCATED_MSG options. By using the BROWSE_FIRST/NEXT option, the message will NOT be deleted from the MQ. After the non-destructive MQGET, we save the value returned in MQMD_MSGID and use the value returned in the specified DataLength field ( rounded up to the next higher 4K boundary ) to GETMAIN enough storage for the WHOLE message. Of course, that means we have to establish addressability to the Getmain'd buffer and define its storage area in the Linkage-Section before it can be referenced. The MQ is then opened a second time and, after storing the saved MQMD_MSGID into the MQMD_CORRELID, a second MQGET is issued with the MQGMO-MATCHOPTIONS set to include MQMO-MATCH-CORREL-ID.
Naturally, GETMAIN's should be matched with FREEMAIN's when you are finished with the buffers. Likewise MQCLOSE's to match MQOPEN's, etc.
_________________
A computer once beat me at chess, but it was no match for me at kick boxing.