Ben,

Monitoring a location that you are directly connected to (either locally or
via a client) is simple since the application need only be able to issue an
API call. When you're going to a remote Queue Manager *via* another then
you actually need to send a message and have it returned. So, yes, going
via another queue manager is more complicated.

I've just tried it though and it still seems to work fine so I recommend
you read the manual and use loopback monitoring. ie. define a remote queue
on your remote QM that 'points' back to the QM your MO71 is actually
connected to.

Cheers,
P.

Paul G Clarke
WebSphere MQ Development
IBM Hursley




|---------+---------------------------->
|         |           "Benjamin F.     |
|         |           Zhou"            |
|         |           <[EMAIL PROTECTED]|
|         |           USA.COM>         |
|         |           Sent by: MQSeries|
|         |           List             |
|         |           <[EMAIL PROTECTED]|
|         |           N.AC.AT>         |
|         |                            |
|         |                            |
|         |           12/03/2004 16:15 |
|         |           Please respond to|
|         |           MQSeries List    |
|---------+---------------------------->
  
>-----------------------------------------------------------------------------------------------------------------------------|
  |                                                                                    
                                         |
  |       To:       [EMAIL PROTECTED]                                                  
                                   |
  |       cc:                                                                          
                                         |
  |       Subject:  Re: MO71 question                                                  
                                         |
  |                                                                                    
                                         |
  |                                                                                    
                                         |
  
>-----------------------------------------------------------------------------------------------------------------------------|



Hi Paul,

I still try to use viaQM because it may take weeks before it can be
installed on the MF.

Just to compare, I added another location whose qmgr is on the same NT
machine. I set it up to be viaQM-monitored. After I double-click the QM,
all the query results are shown on the screen about this QM, its icon turns
green, but just for a short two seconds, it will turn red with 'X' right
away.

What does this mean? Is there more to configure for viaQM monitored QM?

thanks,
Ben






                      Paul Clarke
                      <[EMAIL PROTECTED]         To:
[EMAIL PROTECTED]
                      .IBM.COM>                cc:
                      Sent by:                 Subject: Re: MO71 question
                      MQSeries List
                      <[EMAIL PROTECTED]
                      en.AC.AT>


                      03/11/2004 01:26
                      PM
                      Please respond
                      to MQSeries List






Ben,

If you have the CAF (Client Access Facility) then it begs the question as
to why you're routing your requests to MVS via an AIX Queue manager.
Wouldn't you be better off defining your MO71 MVS location as a client
connection. If you did you would then be able to browse the messages on the
dead letter queue using MO71.

Of course, both solutions require that you either have your userid
registered with your MVS userid or you change the userid of the messages
flowing to MVS to something it will accept.

Cheers,
P.

Paul G Clarke
WebSphere MQ Development
IBM Hursley




|---------+---------------------------->
|         |           "Benjamin F.     |
|         |           Zhou"            |
|         |           <[EMAIL PROTECTED]|
|         |           USA.COM>         |
|         |           Sent by: MQSeries|
|         |           List             |
|         |           <[EMAIL PROTECTED]|
|         |           N.AC.AT>         |
|         |                            |
|         |                            |
|         |           11/03/2004 13:55 |
|         |           Please respond to|
|         |           MQSeries List    |
|---------+---------------------------->

>------------------------------------------------------------------------------------------------------------------------------|


  |
|
  |       To:       [EMAIL PROTECTED]
|
  |       cc:
|
  |       Subject:  Re: MO71 question
|
  |
|
  |
|

>------------------------------------------------------------------------------------------------------------------------------|





Hi Paul,

Many thanks for the hint. I could eventually locate the msg in the deadQ on
mainframe. However, since I can't read msg there as I can on other
platforms, I have to guess what's there. It seems I do have to have CAF to
read the msg. (I'm no MF literate)

I also tried to use Loop-back monitoring, where I make the command queue on
the mainframe a remote queue pointing back to viaQM's reply Queue. Now I
see both channels in running status. Just the MO71 location for the MF qmgr
is still not active. I anticipate the viaQM simply took over the msg before
the MF entry gets a chance to get it. Do you see it that way?

best regards,
Ben






                      Paul Clarke
                      <[EMAIL PROTECTED]         To:
[EMAIL PROTECTED]
                      .IBM.COM>                cc:
                      Sent by:                 Subject: Re: MO71 question
                      MQSeries List
                      <[EMAIL PROTECTED]
                      en.AC.AT>


                      03/10/2004 05:14
                      AM
                      Please respond
                      to MQSeries List






Ben,

Nick has described it perfectly. The whole point of the 'via' location is
that messages are routed to/from your target Queue Manager *via* this other
location.

As far as where did your message go ? Well, check the normal places. If
you're certain the message got to the mainframe then it should have been
processed by the command server.  The most likely thing is that the command
server didn't like your userid (or something) and will have dumped the
request to the dead letter queue.

If all else fails then trace is always an option. With tracing switched on
you should be able to follow the path of any message and see exactly where
it goes. However normally  this shouldn't be necessary.  Check all your
DLQs, command queues and transmission queues and you'll probably find your
message.

Cheers,
P.
Paul G Clarke
WebSphere MQ Development
IBM Hursley




                      Nick Dilauro
                      <[EMAIL PROTECTED]        To:
[EMAIL PROTECTED]
                      >                        cc:
                      Sent by: MQSeries        Subject:  Re: MO71 question
                      List
                      <[EMAIL PROTECTED]
                      N.AC.AT>


                      09/03/2004 21:46
                      Please respond to
                      MQSeries List




Ben,

I'm pretty sure it means that you are using the reply to queue you
specified
when you set up the location for the "via QM".  Since that qmgr has
connections to the other qmgr the replies are routed back to the "via QM".
Even though it doesn't say so, you don't need any of the client info for
the
other qmgr.

Nick


-----Original Message-----
From: MQSeries List [mailto:[EMAIL PROTECTED] On Behalf Of Benjamin
F. Zhou
Sent: Tuesday, March 09, 2004 11:11 AM
To: [EMAIL PROTECTED]
Subject: MO71 question

Hi MO71 veterans,

1. what does it mean that "replyQ is not needed when using via QM". Where
does the reply go to?  to SYSTEM.DEFAULT.MODEL.QUEUE?

Specifically, I'm setting up MO71 to monitor a mainframe QM (without CAF,
I'm now convinced it's not needed) , via a QM on AIX. After I defined
qmgr-alias on the AIX box to point to the MF QM, I see the request got sent
to the mainframe, but no reply ever gets back.  Apparently the MF QM didn't
even try to send a reply back since the XmitQ is empty and sender channel
to AIX never starts.  I also setup qmgr alias on MF pointing to QM on AIX.
The field mvs has been checked.

2. Has anyone ever used the "Command OK" field to run a batch program or
whatever, on a win2k or XP machine?

Paul, maybe you can pin-point what I'm missing right away.

thanks a lot.
Ben

Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://vm.akh-wien.ac.at/MQSeries.archive

Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://vm.akh-wien.ac.at/MQSeries.archive

Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://vm.akh-wien.ac.at/MQSeries.archive

Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://vm.akh-wien.ac.at/MQSeries.archive

Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://vm.akh-wien.ac.at/MQSeries.archive

Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://vm.akh-wien.ac.at/MQSeries.archive

Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://vm.akh-wien.ac.at/MQSeries.archive

Reply via email to