Hi William,
to my mind, MQ clustering or server clustering are methods, to have normal
fail-over or load-balancing features. DR means, to recreate an environment
in a disaster situation (e. g. your whole computer center burns down). In
this case, you need to have a possibility, to recreate your
Rick,
you will have problems, if you use MQ clustering. In this case, different
qmgrs will be seen in the cluster repository as different objects with the
same name. They are different, because of the qmgr IDs (on servers: qmgr
name plus a timestamp). On the other hand they are same - e. g. in
Hi Rebecca,
the saveqmgr program does not store the permissions set with the setmqaut
command. You need another supportpac (I guess MS0I), to save the
authorities too.
Nevertheless, saveqmgr stores only the qmgr configuration, not the qmgr
itself. This means, you will be able, to create a qmgr
I have a CICS transaction that is triggered by the CICS trigger monitor (CKTI).
The local queue attribute fortrigger type is EVERY.
When the transaction is triggered it gets a message off the queue under a syncpoint (MQGMO_SYNCPOINT). It then abends and the message is rolled back to the
Hi Juni,
I am just 2 months old to MQSI. So plz don't take my words for granted.
For Packed Decimal Elements, looks like AS/400 is expecting a different
Data Structure than the one provided by MQSI. When I imported a COBOL
Copy Book with PIC 99v99 COMP-3, for the element COBOL Importer made:
Larry,
you are not allowed to to this in CSQINP1, but CSQINP2 is the right place for
these things...
If it is something that is also needed during the run of mqseries, i would
consider to use
a job for this, which is started after mq startup (by system automation) and
whenever it is needed.
There was an efix for CSD3 on AIX - efix IY39939 - don't know whether it got
included in CSD4 or
not - the 265s were going to the DLQ even though the triggered app was
exiting with rc 0 - the same APAR may exist on the Windows codebase as well.
After the APAR was applied I think you then had to
I have installed and configured QPASA at three client sites. It is a nice
tool and that was prio to Release 3.0 which has many improvements. It allows
you to customize your views of you critical messaging paths (yes MQSI also).
BMC is a nice looking product one client I was at was using it.
Hubert, it depends on the level of MQ you're running -- While earlier
releases required a SupportPac to store the setmqauts (although I just added
them to a file and reran the whole thing when I changed permissions), V5.3
provides a command that will create a file with the required commands. One
SaveQMgr is a good idea
thanks
-Original Message-
From: Bullock, Rebecca (CSC) [mailto:[EMAIL PROTECTED]
Sent: Thursday, September 25, 2003 8:21 PM
To: [EMAIL PROTECTED]
Subject: Re: Disaster Recovery scenario for Unix box?
Bill, if the only reason you're dong the backups nightly
Is the program or transid disabled?
Jeff Horner
[EMAIL PROTECTED] To: [EMAIL PROTECTED]
AIL.COM cc:
Sent by: Subject: MQ v5.3 OS/390 Trigger Problem
Hi,
Here's what I did for DR at my current client. They are new to WMQ and only
have WMQ on Solaris (v5.3).
- Installed SupportPac MS03 - saveqmgr
- Installed SupportPac MS65 - objectSave shell script
- Copied objectSave to oamSave and hacked it to handle security
For WMQ v5.3, amqoamd
I will be out of the office starting September 26, 2003 and will not return
until October 6, 2003.
I will have very limited access to email. For Urgent matter, please contact
my manager Clarence Sham. Thanks!
Instructions for managing your mailing list subscription are provided in
the Listserv
A @#$%$#,
I cut pasted too fast. I forgot the crtmqm task. So here it is (it goes
right above the strmqm!!! grin)
crtmqm -lc -lf 4096 -lp 5 -ls 10 -u SYSTEM.DEAD.LETTER.QUEUE
Regards,
Roger Lacroix
Capitalware Inc.
Quoting Roger Lacroix [EMAIL PROTECTED]:
Hi,
Here's what I
Rick,
The transaction is not disabled.
The program that the transaction executes is not disabled.
I disabled another program that is called viaa COBOL CALL to cause an abend. I needed to test this condition, but it is not working out so well. Here is more details of what is happening in the
No I have to bring up new zata on c
just been too busy
-Original Message-
From: Jeff Horner
[mailto:[EMAIL PROTECTED]
Sent: Friday, September 26, 2003
12:34 PM
To: [EMAIL PROTECTED]
Subject: Re: MQ v5.3 OS/390
Trigger Problem
Rick,
The transaction is not disabled.
Jeff,
Not sure what to say. We do this type of processing all the time without
any problems. It could be a bug. Does it run if you turn triggering off
and on after that?
Jeff Horner
[EMAIL PROTECTED] To: [EMAIL PROTECTED]
Hi Jeff,
If I remember correctly with TRIGTYPE(EVERY), the trigger
message is generated when a message arrives on the queue
however if the transaction abends and the message is backed
out a new trigger message is not generated (that is strictly
from memory and it may have failed me). That would
Hi, have many of you decided to NOT use a Dead Letter Queue? We are
looking into a product that uses MQSeries as a transport layer. The tool
assumes there is no DLQ.
What happens is it sets the Exception Report option so if the remote queue
if full, it returns an exception to the send portion of
Jeff
Why are you expecting a new trigger message? Did you put another message to
the queue?
From my understanding with trigger EVERY, failing to process the message on
the queue can have this result. Basically trigger messages are created as a
result of the put that placed the message on the
Jim is correct. Step 12.b in the trigger conditions specifies that you will
only get retriggered when the message flops back on the queue only for
trigger type of depth or first.
Quote from the manual:
12.
The only application serving a queue issues an MQCLOSE call, for a
TriggerType of
We've been through this before. I don't see how anyone can run a large
scale MQSeries infrastructure without the use of a dead letter queue. If a
single queue becomes full or a single user is not authorized to a remote
queue, the channel will be stopped. This means that any other application
There is no way I'm going to sacrifice my channels by not having a DLQ,
when some nondeliverable message comes across the channel without a
disposition value of MQRO_DISCARD_MSG. My sleep at night is to important to
me and the health of the channels is far to important to the rest of the
business
23 matches
Mail list logo
