ASG
IBM
Zystems
Cressida
Icon
Netflexity
 
  MQSeries.net
Search  Search       Tech Exchange      Education      Certifications      Library      Info Center      SupportPacs      LinkedIn  Search  Search                                                                   FAQ  FAQ   Usergroups  Usergroups
 
Register  ::  Log in Log in to check your private messages
 
RSS Feed - WebSphere MQ Support RSS Feed - Message Broker Support

MQSeries.net Forum Index » General IBM MQ Support » channel communications failure : AMQ9213

Post new topic  Reply to topic
 channel communications failure : AMQ9213 « View previous topic :: View next topic » 
Author Message
Tim
PostPosted: Fri Oct 14, 2005 5:49 am    Post subject: channel communications failure : AMQ9213 Reply with quote

Newbie

Joined: 14 Oct 2005
Posts: 4

Hi,

I have found similar problems on this form but unfortunately without a documented solution. Here 's our problem :
we communicate with our customer using MQseries, with a QMGR on both sides. Our customer uses W2000, we run on AIX 5.2, both running MQ version 5.3

This runs very well since over 1 year, untill suddenly we got communication errors earlier this week.
The messages remained in the customer 's transmission queue, and the channels failed communicating. On TCP level I saw established connections again and again, and when sniffing the network the messages were transmitted up to our MQ server. The firewall log shows no drops, in fact we see that the channels attempt reconnecting every 6 minutes. After a few hours the problem was suddenly solved, neither the client or we changed anything (??).

Today the problem came back, communications failure. ..
Our MQ error log on AIX shows this :

----- amqrccca.c : 804 --------------------------------------------------------
14/10/05 15:37:28
AMQ9213: A communications error for TCP/IP occurred.

EXPLANATION:
An unexpected error occurred in communications.
ACTION:
The return code from the TCP/IP(select) [TIMEOUT] 360 seconds call was 11
(X'B'). Record these values and tell the systems administrator.

Error x'B' = 11 means 'reource not available'.
Means nothing to me. Server has plenty of resources, none of the system limits have exceeded I think, CPU ok, paging space ok, errorlog ok.

The listener is respawned from the inetd. So I tried to start it with runmqlsr instead, but the problem remains...

Our network setup is like this :
iVPN connection, firewall on customer side and on our side.
Lines are stable, firewall ok on both sides, TCP communications on our channel port can be established (telnet ok)

I have stopped all QMGRS on our server, remove hanging shared memory and semaphores, and started up our QMGRS agains.
All QMGRS run fine again, all customers connect without problems except the same customer, the problem remains. Customer called IBM support, who was apparantly convinced that the problem was on our side.

In fact I did everything except rebooting the box.

Has anybody any idea ?

Thanks in advance.
Back to top
View user's profile Send private message Send e-mail
Nigelg
PostPosted: Fri Oct 14, 2005 6:29 am    Post subject: Reply with quote

Grand Master

Joined: 02 Aug 2004
Posts: 1046

This is an error in the TCP stack.

The error from select() just means that the socket did not become available for either writing or reading (depending on what operation was being attempted) during the 360 seconds that WMQ waited for it. The error 11 does not mean anything there, the real point is that the select never popped with return code 0 to say that the socket was now ready for the requested operation. The channel then had no option but to end.

You have to find the problem in TCP or the network somewhere.

your greatest obstacle here is getting your network guys to admit that there might be something wrong.
_________________
MQSeries.net helps those who help themselves..
Back to top
View user's profile Send private message
Tim
PostPosted: Fri Oct 14, 2005 7:10 am    Post subject: Reply with quote

Newbie

Joined: 14 Oct 2005
Posts: 4

Thank you gransmaster Nigel. We see lots of TCP retranmissions on this network link so you are probably right.
This doesn't make the situation easier of course because part of the network is managed by the customer, part is managed by us. Our part seems to be perfectly fine dixit the network gurus.
Still it 's amazing than "ping channel" always works, and communication is established "netstat -a", and telnet on the MQ listener port works fine also.
Meanwhile the problem was resolved automaticall, just like earlier this week. Could you recommand some TCP parameter to verify or change ? They are all pretty standard, maybe a MQ of WAN requires some tuning ?
Back to top
View user's profile Send private message Send e-mail
Tim
PostPosted: Tue Oct 18, 2005 1:45 am    Post subject: Reply with quote

Newbie

Joined: 14 Oct 2005
Posts: 4

Problem solved.
Our network engineers found no problem on any of the communication HOPS, but after rebooting a VPN box the problem was definitely solved.
It was hard to identify this as a communication trouble because there are a few firewalls in between, which did not allow TCP traffic other then MQ.
Thanks for help.
Back to top
View user's profile Send private message Send e-mail
Display posts from previous:   
Post new topic  Reply to topic Page 1 of 1

MQSeries.net Forum Index » General IBM MQ Support » channel communications failure : AMQ9213
Jump to:  



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
Protected by Anti-Spam ACP
 
 


Theme by Dustin Baccetti
Powered by phpBB © 2001, 2002 phpBB Group

Copyright © MQSeries.net. All rights reserved.