| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| MQ Channels |
« View previous topic :: View next topic » |
| Author |
Message
|
| rammer |
 Posted: Fri Oct 12, 2012 2:19 am Post subject: MQ Channels |
 |
|
Partisan
Joined: 02 May 2002
Posts: 359
Location: England
|
Morning All. Quick question and I have read the relevant document over and over again but have a mental blockage on understanding it.
Scenario MQ 6.0.2.10 running on Windows
Maximum channels is set to 350
Application is connecting using SVRCONN Channels.
This is regularly hitting 350 and allowing no more connections. I believe that the application has ended the conn's and I am told be the developers this has been done cleanly. the application is sending a request and getting a response back and normally is done within a few seconds.
The settings I have for the channels is the following
Maximum Initiators 3
MQI Bind type standard
Adopt new MCA ALL
Adopt new MCA check: NAME, ADDRESS, QM
Adopt new MCA timeout: 60
Pipeline length: 1
I cant get my head clearly around how the adopt new MCA and timeout works, if this needs changing.
Thanks in advance |
|
| Back to top |
|
 |
| mqjeff |
| Posted: Fri Oct 12, 2012 3:58 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
It's generally a bad idea for an app to connect and disconnect for each request/response operation.
MQ v6 is out of support, you should at a minimum upgrade the queue manager to a supported level.
In your configuration, what you have said is that if a new connection attempt is made for a specific channel from a specific address and there is an existing channel agent already running, then the code will wait up to 60 seconds before killing that agent and acquiring a new one.
You probably need to look at KeepAlive, rather than AdoptNewMCA.
If the app process itself isn't closing down, but merely closing and opening connections on multiple threads, there is a large potential for behind the scenes caching of connection handles.
And even if you can't upgrade the queue manager to a supported level, you may find more reasonable behavior if you upgrade the client install to a newer level. |
|
| Back to top |
|
|
| rammer |
| Posted: Fri Oct 12, 2012 4:36 am Post subject: |
|
|
Partisan
Joined: 02 May 2002
Posts: 359
Location: England
|
| Thanks for the reply. At the moment were not in position to upgrade to 7 for this server but do have extended support. Keepalive is set to Yes. I just need to have a play around and more importantly look at what the application is actually doing For some reason (blame last nights outing) I couldnt get my head around the MCA part!.. |
|
| Back to top |
|
|
| mqjeff |
| Posted: Fri Oct 12, 2012 5:26 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
| KeepAlive is enabled? That's good. You might want to lower the keepalive interval in the OS, then. |
|
| Back to top |
|
|
| rammer |
| Posted: Thu Nov 01, 2012 4:08 am Post subject: |
|
|
Partisan
Joined: 02 May 2002
Posts: 359
Location: England
|
Hi mqjeff apologise for the delay Im guessing im loking in the wrong area but the following was a suggestion by IBM when we had support for 6, which we are now looking to upgrade. Not all the errors are due to us hitting high volume of SVRCONN Channels.
Windows 2003
Settings we have are
Maximum Initiators 3
MQI Bind type standard
Adopt new MCA ALL
Adopt new MCA check: NAME, ADDRESS, QM
Adopt new MCA timeout: 60
Pipeline length: 1
TCP KeepAlive Yes
TCP Listener backlog 100
Within the Windows Reg I have the following
HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\parameters\
KeepAliveInterval = 1000
KeepAliveTime = 300000
TcpMaxDataRetransmissions=5
As per document http://www-01.ibm.com/support/docview.wss?uid=swg21190501
The application is local to the Server coming in via IIS here is an example of the errors we are seeing
01/11/2012 11:34:44 - Process(5524.3804) User(svcwwwPrtLabel) Program(w3wp.exe)
AMQ9202: Remote host '172.20.160.213(1414)' not available, retry later.
EXPLANATION:
The attempt to allocate a conversation using TCP/IP to host
'x.x.x.x(1414)' was not successful. However the error may be a
transitory one and it may be possible to successfully allocate a TCP/IP
conversation later.
ACTION:
Try the connection again later. If the failure persists, record the error
values and contact your systems administrator. The return code from TCP/IP is
10061 (X'274D'). The reason for the failure may be that this host cannot reach
the destination host. It may also be possible that the listening program at
host 'x.x.x.x(1414)' was not running. If this is the case, perform the
relevant operations to start the TCP/IP listening program, and try again.
and not quiet as many of these but still alot
Process(6268.134716) User(MUSR_MQADMIN) Program(amqrmppa.exe)
AMQ9208: Error on receive from host xxxx (x.x.x.x).
EXPLANATION:
An error occurred receiving data from xxxx (x.x.x.x) over
TCP/IP. This may be due to a communications failure.
ACTION:
The return code from the TCP/IP (recv) call was 10054 (X'2746'). Record these
values and tell the systems administrator.
----- amqccita.c : 3383 -------------------------------------- |
|
| Back to top |
|
|
| mqjeff |
| Posted: Thu Nov 01, 2012 4:54 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
This is all related to the fact that the app is connecting and disconnecting frequently.
Those errors mean that keepalive or something else is killing the channel.
Windows 10054 message is ECONNRESET - something has killed the tcp connection. |
|
| Back to top |
|
|
|
|
|
| |
|
Page 1 of 1 |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
