users@glassfish.java.net

Re: Glassfish v2.1 JMS stop calling consumers

From: Paulo Cesar Reis <casmeiron_at_gmail.com>
Date: Wed, 07 Oct 2009 12:18:34 -0300

Hi Nigel,

The problem happened again, in other machine on the same network. This time
I had a chance to execute the command on imq, follow:
Listing all the destinations on the broker specified by:

-------------------------
Host Primary Port
-------------------------
localhost 7676

----------------------------------------------------------------------------
--------
         Name Type State Producers Consumers Msgs
                                                     Total Count UnAck
Avg Size
----------------------------------------------------------------------------
--------
NPAlarmHandlerQueue Queue RUNNING 0 1 0 0
0.0
NPAlarmNotifierTopic Topic RUNNING 0 1 0 0
0.0
NPEventChannelTopic Topic RUNNING 0 2 0 0
0.0
NPMessageReceiverTopic Topic RUNNING 0 1 3811 559
1361.94
NPMessageReplyQueue Queue RUNNING 0 1 1 1
379.0
mq.sys.dmq Queue RUNNING 0 0 0 0
0.0

Successfully listed destinations.


As you can see, my Topic " NPMessageReceiverTopic" has tons of messages that
is never bein' delivered, the JMS just stop calling consumers but continue
to accept new messages, I hope you can help me.

Thanks.


On 10/6/09 3:42 PM, "Paulo Cesar Silva Reis" <casmeiron_at_gmail.com> wrote:

> Hi Nigel,
>
> Thanks again for your dedication.
>
> The weird problem is that I just have one server in production per client,
> when I said I got 12 servers they're all in different clients, different
> networks, we do not use any kind of cluster on mqbroker nor glassfish. I'm
> using topic but currently I have only one MDB deployed that is able to
> process this kind of message.
>
> I'll try to turn on the problematic server again to execute the given
> command maybe we could have some clue about what's goin' on but for me the
> JMS Broker just stop calling consumers and if I try to stop glassfish (with
> asadmin stop-service) the broker don't stop automatically, I have to kill it
> (yeah). We're also making some tests on RAM and HD to check whether any
> corruptions exist. As soon I get the results of the command ill report you
> back.
> Thanks.
>
>
>
> On 10/6/09 1:16 PM, "Nigel Deakin" <Nigel.Deakin_at_Sun.COM> wrote:
>
>> and
>
>