| Author |
Message
|
| mjoshi |
 Posted: Tue Mar 27, 2007 6:39 am Post subject: AMQ9513: Maximum number of channels reached |
 |
|
Newbie
Joined: 27 Mar 2007
Posts: 8
|
I am getting channels maxed out even when bumping the # of channels to 1000. Below is my configuration file(qm.ini) and the error. NO FDC's generated
Error
------
ExitPath:
ExitsDefaultPath=/var/mqm/exits/
#* *#
#* *#
Log:
LogPrimaryFiles=10
LogSecondaryFiles=4
LogFilePages=16384
LogType=CIRCULAR
LogBufferPages=512
LogPath=/var/mqm/log/XXX_XXX_XXX/
LogWriteIntegrity=TripleWrite
Service:
Name=AuthorizationService
EntryPoints=10
ServiceComponent:
Service=AuthorizationService
Name=MQSeries.UNIX.auth.service
Module=/opt/mqm/lib/amqzfu
ComponentDataSize=0
Channels:
MaxChannels=1000
MaxActiveChannels=1000
TuningParameters:
DefaultPQBufferSize=1000000
Error
------
AMQ9513: Maximum number of channels reached.
EXPLANATION:
The maximum number of channels that can be in use simultaneously has been
reached. The number of permitted channels is a configurable parameter in the
queue manager configuration file.
ACTION:
Wait for some of the operating channels to close. Retry the operation when some
channels are available.
----- amqrcsia.c : 1026 -------------------------------------------------------
03/27/2007 10:02:19 AM
AMQ9999: Channel program ended abnormally.
EXPLANATION:
Channel program 'SYSTEM.DEF.SVRCONN' ended abnormally.
ACTION:
Look at previous error messages for channel program 'SYSTEM.DEF.SVRCONN' in the
error files to determine the cause of the failure. |
|
| Back to top |
|
 |
| Michael Dag |
| Posted: Tue Mar 27, 2007 6:45 am Post subject: |
|
|
Jedi Knight
Joined: 13 Jun 2002
Posts: 2607
Location: The Netherlands (Amsterdam)
|
|
| Back to top |
|
|
| mjoshi |
| Posted: Tue Mar 27, 2007 7:05 am Post subject: |
|
|
Newbie
Joined: 27 Mar 2007
Posts: 8
|
| We have 30 clients conencting to the Qs. Also, a "netstat -an | grep 1414 | wc -l" yields 1657 conections. I don't know if that helps. |
|
| Back to top |
|
|
| Vitor |
| Posted: Tue Mar 27, 2007 7:09 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| mjoshi wrote: |
| We have 30 clients conencting to the Qs. Also, a "netstat -an | grep 1414 | wc -l" yields 1657 conections. I don't know if that helps. |
If I had 30 clients, and 1,657 connections I would start to wonder if all the clients were disconnecting properly. Especially if I used to have a limit of 100 connections which was exceeded so I upped the limit to 1000. If the limit was put up again (say 1500) and that limit was exceeded, but still with the same 30 clients I would wonder seriously about that... 
Are you using an application server or similar connection pooling mechanism? Anything which might provide a legitimate explaination for the high number of connections?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| mjoshi |
| Posted: Tue Mar 27, 2007 7:17 am Post subject: |
|
|
Newbie
Joined: 27 Mar 2007
Posts: 8
|
| Yes, I am using Websphere application server. Then there are job servers which do most of the messaging. This ran fine in the dev envioronment where we were running the agents and MQ on the same box. However, now we moved to "staging" which is a distributed setup, I can't help but wonder if this is purely a network connectivity issue. |
|
| Back to top |
|
|
| Vitor |
| Posted: Tue Mar 27, 2007 7:32 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
If the agents run on these job servers do these all use pooled connections under the app server? If so, what's the syncronisation (if any) between app servers and queue manager?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| mjoshi |
| Posted: Tue Mar 27, 2007 7:48 am Post subject: |
|
|
Newbie
Joined: 27 Mar 2007
Posts: 8
|
| No, I don't think so. We are not defining any bindings on the application server. |
|
| Back to top |
|
|
| tleichen |
| Posted: Tue Mar 27, 2007 7:51 am Post subject: |
|
|
Yatiri
Joined: 11 Apr 2005
Posts: 663
Location: Center of the USA
|
| mjoshi wrote: |
| ... This ran fine in the dev envioronment where we were running the agents and MQ on the same box... |
Yes, but typically in dev, you don't have the number of clients that you do in production and/or you do not run the scenario for as long of a time. Therefore, even if your client apps are not doing things properly, the problem may not manifest itself in that environment.
_________________
IBM Certified MQSeries Specialist
IBM Certified MQSeries Developer |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Tue Mar 27, 2007 9:17 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
Hey, Mayank. 
You may need to increase the number of connections. Your agent processes are using JMS, and typically JMS uses significantly more connections than people expect them to.
For example, for a single instance of an MDB, there will be at least two connections - one used by the listener port and one used by the MDB itself. The listener port will be scanning the queue constantly for new messages, and then alert the MDB that a new message has arrived, and the MDB will then open a connection to get the message.
You can try and list connections, and associate them with a particular agent process, and see if you can establish some estimates for how many connections you actually need. Also, given that your different agent processes are doing different things, there may be more connections needed for one process than for another.
This is going to be easier to do if you've migrated to MQ v6 than it is for v5.
I would hope that your software vendor for this application would be able to help you characterize this, as well.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
|
|
