Peter,
i get a little bit confused by the mix of your remotequeues and
qmgr aliases, and the setup you described in this mail sounds
different from the setup you described before, but both
setups look like a works as designed to me.
see my comments () below.
regards
stefan
Title: RE: Problem with MQAX200.dll
Reddy:
The MQQueue class has a property CurrentDepth which will return a long containing the info. you want.
Cheers,
Andy.
-Original Message-
From: vemulapati narasimha reddy [mailto:[EMAIL PROTECTED]]
Sent: 21 November 2002 18:52
To:
I am no COM expert, but looking into the MQ Series information centre this
is what is mentioned
This error occurs while doing an MQOPEN without specifying an object type.
2043 - MQRC_OBJECT_TYPE_ERROR
On the MQOPEN or MQPUT1 call, the
Maybe I don't see your point, but doesn't that negate the benefits of
Symmetrical Multi-Processing? Don't do for the operating system what the
operating system will do for you.
-Original Message-
From: MQSeries List [mailto:[EMAIL PROTECTED]]On Behalf Of Potkay,
Peter M (PLC, IT)
Sent:
When MQSI is doing its thing and is real busy with heavy volume going thru
an ESQL heavy compute node, the MCAs on the box get left with no resources.
As a result we see XMIT queues to this box temporarily back up as the
sending MCAs can't talk to the receivers, because the receivers have no
OH!!! NO!! It's the LISTSERVer Police!
bee-oh-dubble-bee-dubble-egh!!
From: Christopher Fryett [EMAIL PROTECTED]
Reply-To: MQSeries List [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject: Re: Websphere MQ questions
Date: Sat, 23 Nov 2002 09:51:20 -0600
Tia,
Peter, I think we are experiencing this problem also with the sending MCA
on our NT qmgr/broker
machines.
I thought I had the same problem with the receiving MCA, although it turned
out to be security related. For instance
we have messages that are sent from an OS/390 qmgr to our NT qmgr/broker
Because of the sporadic volume of messages, and the fact there are
applications that need their messages ASAP, we have the BATCHINT set to
zero. we need the message to go over the wire right away.
I don't know if running the MCA on the MQSI queue manager as a FASTPATH app
would help or not, or
What version of MQ Series is required on MVS to allow message segmentation?
Anne Hoshooley
Systems Specialist - Information Services
ATT Canada
(416)345-3486
on net 8-862-3486
teleworking (519)772-3008
Instructions for managing your mailing list subscription are provided in
the Listserv General
If I change a queue to persistent, does
that change take effect immediately or do I have to restart the QMGR (NT5.2).
Thanks
Can it be connected with the length of the message?
Ping message is short.
Eugene Rosenberg
--- Pope, Ben [EMAIL PROTECTED] wrote:
I've got a situation between an Win2K (5.3) and
AS/400 (5.2) box with a sender channel stuck
in retrying mode. If I manually stop the channel I
can ping it
Dave,
The change is immediate for all new messages put directly into the queue by
the application, where the application does not specify anything on the
put.
so if you are using and alias, and change only the base, that does not
change any messages put via the alias. conversely if you change
Title: Exporting a flow from WMQI for promotion/version control purposes.
When I export a flow from WMQI that uses sub-flows, the dependant sub-flow(s) get exported into the file also. Is there any way I can export each flow to a separate file without each export containing all the dependant
No version of MQ for OS/390 or z/OS supports segmentation, though version 5
does support conversion of the MDE, so you can do you own segment
reassembly.
Dave
Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at
Keep in mind if you don't have the required version, you can still break
apart messages and send to MVS. It would be an app-specific algorithm, but
between get-by-correl id and knowing how many messages made up the unit of
work, it's not too difficult to code yourself. Also could use AMI send /
Checking the suggested logs, it was a sequencing error. Since I reset
the channels its been working fine.
I am curious though as to what types of situations can give rise to a sequencing
error. In production (if we ever get that far) the systems will be operating
across a VPN with
Does anyone have a good knowledge of Linux programming with MQSeries?
I am trying to recompile some of my C++ programs on RedHat Linux 8.0.
The MQSeries (5.3.1) readme says I need libstdc++.so.3 to compile the
samples (which my installation didn't have), so I dutifully downloaded
an rpm package
Ben,
Once the channels are running and working fine you should not get any
sequencing errors, unless some one actually tampers with the syncq or manually
issues reset channel. The network problems should not effect the seq number. So
the chances for you to get similar errors in production is
Title: Checking the MQMD Expiry in WMQI
If a message is set to unlimited expiry, the MQMD.Expiry is treated as an integer in WMQI. If the message is not set to unlimited expiry the MQMD.Expiry is treated as a timestamp.
If I don't know what the Expiry interval is, how do I compare it without
Date sent: Mon, 25 Nov 2002 12:49:49 -0500
Send reply to: MQSeries List [EMAIL PROTECTED]
From: Pope, Ben [EMAIL PROTECTED]
Subject:Re: Retrying Channel Status?
To: [EMAIL PROTECTED]
Checking the suggested logs, it was a
Hi Andrew
Thank you very much for the reply but ..I have tried to get depth value using currentDepth property type..But it's giving me error like "Option are wrong.."
and MQQueue:.GetCurrentdepth:""
What type options I have to set for MQQueue and for MQGetMessageoptions..
regards
/Reddy
/Reddy
We have an application which at times will be processing lots of messages
in a short period of
time. Somewhere around 100,000 request per hour, followed by the same
reply message amounts.
Then another 100,000 of larger messages flowing in the same direction of
the reply.
So a total of 300,000 per
hat are your error logs saying??//
From: eugene rosenberg [EMAIL PROTECTED]
Reply-To: MQSeries List [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject: Re: Retrying Channel Status?
Date: Mon, 25 Nov 2002 08:41:18 -0800
Can it be connected with the length of the message?
Ping
I have lost hem again!!
I need the actual codes for the Channel Status (ie 1,2,3,4,5,6) and their
meaning. I looked in the Admin, Tele and Pgmrs Ref. I am either missing it
or IBM pulled them out. I get the inactive, binding, running' stuff but not
the codes for the English. I'm trying to set up
If the messages are non-persistent, then the impact on disk I/O should be
minimum. Non-persistent messages are kept in buffers and only go to real
disk when buffer space is exhausted. As long as the applications pull the
messages at a good clip, the over-flow to real disk should be a small risk.
My list gives me:
0 INACTIVE
1 BINDING
2 STARTING
3 RUNNING
4 STOPPING
5 RETRYING
6 STOPPED
7 REQUESTING
8 PAUSED
13 INITIALISING
HTH
John.
-Original Message-
From: Robert Broderick [mailto:[EMAIL PROTECTED]]
Sent: 25 November 2002 18:56
To: [EMAIL PROTECTED]
Subject: Channel
Title: Checking the MQMD Expiry in WMQI
Hi
John,
Can
you tell me where this behavior is documented?
Thanks
Karen
-Original Message-From: John Scott
[mailto:[EMAIL PROTECTED]]Sent: Monday, November 25, 2002 1:16
PMTo: [EMAIL PROTECTED]Subject: Checking the MQMD
Expiry in WMQI
If a
Reddy,
I'm not terribly familiar with MQAX but did you include the MQOO_INQUIRE
option on your queue open call?
HTH,
Marty Frasier
e-mail: [EMAIL PROTECTED]
Narasimha reddy.V [EMAIL PROTECTED]
Sent by: MQSeries List [EMAIL PROTECTED]
11/25/2002 01:27 PM
Please respond to MQSeries List
I'm not an expert in C, and there is probably a better way of writing this,
but this works for me (similar to your assignment statement):
strncpy(od.ObjectName,queue,MQ_Q_NAME_LENGTH);
where od is declared previously with:
MQOD od = {MQOD_DEFAULT};/* Object Descriptor */
and
Got MQ V5.3 on NT and seem to be having some problems getting clustering to
work properly. I joined an NT Qmgr to a cluster and if I do a dis
clusqmgr(*) in any of the other AIX Qmgrs I can see the joined NT Qmgr but
I can't see any of the shared queues. Nor can I PUT to the shared queues (of
Channel Initiation on OS/390. We are OS/390 V2.8 amd MQ Series V2.1.
MQ on OS/390 has been configured with the appropriate Queue and Channel
definitions. However, channels communicating with distributed MQ Qmgrs
refuse to initiate. The XmitQ definition is set with a Process that is
defined with
Two things - the process definition must have the channel name in the
USERDATA field
second - I'm not totally sure the applid is QX52 if you are on MQ 2.1,
unless you are being pro-active and just renamed the old CSQX module (thats
the applid we point to for 2.1).
Peter
From: Mark Steely
If there are messages on a queue that is set to trigger a CICS transaction
on FIRST, and the CICS region is down, will a trigger message be generated
when the CICS region and CKTI trigger monitor first come up? I thought yes,
but IBM support just told me no. I don't want to spend a lot of time
Bruce,
Have you tried explicitly defining a Cluster Sender queue from the
repository back to the NT box? I had similar problems until I read that the
repository information flows along explicitly defined channels. Once I did
that, the cluster on my NT box refreshed itself of it's own accord.
Mark, I assume that it's not something as simple as remembering to turn on
the triggering?
We're running V5.2, but I don't think that should make a difference on this.
On our initiated channels, the process definition says CSQX START. Where
did the QX52 come from? Or was it a typo?
-- Rebecca
Was that a typo for the APPLICID ?
Ours have: Application ID : CSQX START. The command Ref manulal says
that's what it should be for distributed queueing without CICS.
Richard.
Mark Steely [EMAIL PROTECTED]@AKH-WIEN.AC.AT on 26/11/2002 09:25:39
Please respond to MQSeries List [EMAIL
char queue[49]; /* Queue name*/
queue[49] = '\0';
I think you really meant the following.
queue[48] = '\0';
-Original Message-
From: MQSeries List [mailto:[EMAIL PROTECTED]]On Behalf Of Peter
Heggie
Sent: Monday, November 25, 2002 2:00 PM
To: [EMAIL
Title: MQe usage
Gary,
We
used MQe (version 1.26) to support a customer's need to support about 1000 field
engineers who use laptop devices (some Win98 some Win2K) to obtain
information from an MQSeries enabled set of legacy applications. The key
was designing the MQe topology, and the
You need to have the sender channel name in the USERDATA field of the
process. Also, I believe the APPLICID parameter has to be CSQX START (not
QX52, which I assume is your queue manager name). And the channel initiator
has to be running, of course.
-Original Message-
From: Mark
Peter,
If you quote a literal with double quotes, the C compiler automatically
null-terminates the string for you. Therefore, mqm results in a
4-character array of 'm', 'q', 'm', '\0'. The single-quotes denote a
character constant which is a single character and not automatically
Yes it will.
If you do not see it in test then look at your initq.
Dave
-Original Message-
From: MCSHEFFREY, MICHELLE (AIT) [mailto:[EMAIL PROTECTED]]
Sent: Monday, November 25, 2002 3:38 PM
To: [EMAIL PROTECTED]
Subject: Triggering when CKTI comes up
If there are messages on a queue
One of the times that the QM checks to see if a trigger message should be
generated is when the INIT queue IPROCS count goes from 0 to 1.
I would think that when CKT1 is brought up, it opens the INIT queue, and if
no other app had the init queue open, the IPROCS would then go from 0 to 1.
The
Make sure you authorize yourself for the
SYSTEM.CLUSTER.TRANSMIT.QUEUE for the cluster
queue manager. This has bitten me a couple of
times. Normally you would expect a 2035 back
but I remember getting some weird error back
and opening a PMR with IBM. The resolution was
to set the authorities on
Thatnk John. What manual did you find them in?
bobbee
From: John Scott [EMAIL PROTECTED]
Reply-To: MQSeries List [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject: Re: Channel Status Codes
Date: Mon, 25 Nov 2002 19:04:53 -
My list gives me:
0 INACTIVE
1
One of the prerequisites for trigging is that a trigger monitor must have
the initiation queue open for input. This is likely the reason.
-Original Message-
From: MCSHEFFREY, MICHELLE (AIT) [mailto:[EMAIL PROTECTED]]
Sent: Monday, November 25, 2002 12:38 PM
To: [EMAIL PROTECTED]
Unless IBM changed things and I may be mistaken, I have been away from OS390
for a year and a half so the finer details are getting fuzzy. When you
restart a CICS Trigger Monitor the triggers get fired again. Anybody
else
bobbee
From: MCSHEFFREY, MICHELLE
Thanks Steve and Peter.
I used the strcpy and bang! it worked.
Tomorrow, I start playing with the code to see what I can control (secure).
_
Regards,
Dick Killian
MQ Administrator Adabas DBA
Rochester Gas Electric Corp
(585) 771-6049
GIES,
It that CSQX with a space then START or all 1 word? Is there anything
that I need to do to get CSQXSTART setup? I can't find this in my
manuals.
Thank You
[EMAIL PROTECTED] 11/25/02 02:42PM
Mark, I assume that it's not something as simple as remembering to turn
on
the triggering?
We're
If you down load and install supportpac MA10 for ISPF,
it comes with deadletter queue option.
try this site :
http://www-3.ibm.com/software/ts/mqseries/txppacs/ma10.html
--- Mark Steely [EMAIL PROTECTED] wrote:
We are OS/390 v2.8 and MQ Series V2.1. Does any one
have a dead letter
queue
The person setting this up was copying an old definition that never
worked.
[EMAIL PROTECTED] 11/25/02 02:30PM
Was that a typo for the APPLICID ?
Ours have: Application ID : CSQX START. The command Ref manulal says
that's what it should be for distributed queueing without CICS.
Richard.
Bobbeee,
take a look at the include file cmqcfc.h:
/* Channel Status */
#define MQCHS_INACTIVE 0L
#define MQCHS_BINDING 1L
#define MQCHS_STARTING 2L
#define MQCHS_RUNNING 3L
#define MQCHS_STOPPING 4L
#define MQCHS_RETRYING 5L
#define MQCHS_STOPPED 6L
#define
1) We are running V2.1 broker on Solaris8. How do I tell what database a
broker is attached to and what the database userid was set to AFTER an
mqsicreatebroker has been done?
2) Is the database that is accessed in a compute the same database that is
set in the mqsicreatebroker command -n
Mark,
it is CSQX START (with a blank). For reference to documentation, please take
a look at the MQSeries Command Reference
http://www-3.ibm.com/software/ts/mqseries/library/manualsa/csqzaj/csqzaj0y.htm
and scroll down to the description of the APPLICID parameter.
Hope that helps,
Stefan
From:
CSQX (space) START
Fred Oddo
Securities Industry Automation Corporation
CICS / MQ Systems Programmer
(212) 383 4087
Mark Steely
Mark.Steely@WNTo: [EMAIL PROTECTED]
CO.COMcc:
Sent by:
Bob -
these codes and codes for channel types can be found in java pcf support
pack manual. I was just looking them up myself couple days ago. Those codes
are actually supported by qpasa, so you can use them for event monitoring
definitions.
Mikhail.
We have MA10 pac and it only allows you to delete the messages. We would
like to re-load the messages. Any help would be appreciated.
Thank You
[EMAIL PROTECTED] 11/25/02 03:45PM
If you down load and install supportpac MA10 for ISPF,
it comes with deadletter queue option.
try this site :
Thank for all your responses. The process def was in error.
Thank You
[EMAIL PROTECTED] 11/25/02 02:42PM
Mark, I assume that it's not something as simple as remembering to turn
on
the triggering?
We're running V5.2, but I don't think that should make a difference on
this.
On our initiated
As far as I know you cant set processor affinity under NT however you can
set priorities for processes and tasks. You may want to try setting you MQ
processes with a higher priority than MQSI(WMQI). After all it is WMQ that
is servicing WMQI and as such you want it receiving priority.
Regards
Tim
Srivathsa.
Thank you for your reply, but I am afraid this is not the right track:
as I said, everything works if I do not specify a message length in the Get.
Unless I learn that the addition of this parameter leads to another
queue being opened (which seems to be the case but I cannot
I agree with the blocking concept. However for your current problem if you
are using circular logging you are probably running out of log file space
so I'd suggest you increase the number of log files you are using. The
other parameter you may encounter trouble with is MAXUMSGS i.e. the number
of
Simple way to look at it. MQ provides services to CICS so if CICS is
running at a higher priority then it keep pre-empting its service provider.
Regards
Tim A
Hi Listers,
I got the amq6004 problem in the SUN box this morning which caused the queue
manager cannot be brought up. This is the second time it happened on that
box this year and thanks for those who showed me the amqiclen command
recently on the list. This time I don't have to issue ipcs and
The messages themselves may not be persistent but I'm guessing that the
receiving application does something with all that data, thus whilst MQ may
be able to manage them in memory there will still be a significant I/O
requirement by the app.
Regards
Tim A
Malamud,
I think it also depends on the trigger interval. Once the CKTI is up and
running, the trigger message will be generated within the trigger interval
period.
Ian
-Original Message-
From: Potkay, Peter M (PLC, IT) [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, 26 November 2002 8:02 AM
To: [EMAIL
Someone also put some useful MQ documents in Palm format including that
channel status below, reason code, etc. If you have a Palm, it's quite
handy and worth for download.
See http://www.qvadis.com/exlibris/library-computers.html
** I've never verified the correctness but the reason code I've
What options do you use for endmqm command?
-Original Message-
From: Chan, Ian M
To: [EMAIL PROTECTED]
Sent: 11/25/2002 7:31 PM
Subject: amqiclen
Hi Listers,
I got the amq6004 problem in the SUN box this morning which caused the
queue
manager cannot be brought up. This is the second
It's a normal shutdown (i.e. the default control shutdown). Actually the
shutdown worked without problem. I can see the queue manager ended messsage
in the message log.
Ian
-Original Message-
From: Malamud, Mikhail [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, 26 November 2002 1:12 PM
To:
The first thing that comes to mind is
your queue manager is not completely shutdown and then you try to start
a half shut down queue manager. This can happen if your script is
not smart enough to deal with a slow or failed shutdown. Misbehaving
apps or apps not designed to properly recognize the
A Sun Blade is a workstation
so that is the first concern. These also come in various sizes, number
of CPU, memory and so forth so this is really not enough information. You
have to consider all the resources and application design to size your
server. You 300,000 per hour is really nothing to an
Is it correct to state that MQSeries 2.1 supports message conversion ( codepage conversion) on the MQGET
i.e. supports MQPMO_CONVERT.
I was under the impression that it did not or perhaps previous versions did not.
Cheers!
Kevin
Kevin Tobin
Senior IT Specialist
IBM WebSphere Services Asia
We
havea similar arrangement except we stop at -i step. Actually in the
case happened this time, all the processes are ended and the shutdown took 1min
6 sec to complete. The queue manager was then brought up 15 minutes later by the
script but failed with amq6004. When I got paged, I grep the
amqiclen is the IPCS clean up. If
you have this problem, it is usually caused by applications refusing to
let go of the queue manager and that don't recognize when a queue manager
is stopped. It can just happen too, but if you close the applications
first before stopping the queue manager, you
It supports MQGMO_CONVERT, but not MQPMO_CONVERT
:) I know this was a typo, but just to be sure..
I know for sure that 2.1 supports it
and if memory serves, so does 1.2, which is the oldest on OS/390 I have
been exposed to.
Mike Murphy
Sr. Middleware Consultant
MQ Solutions, LLC
73 matches
Mail list logo
