Re: Negtive MessageCountAwaitingAcknowledge & DispatchedQueueSize on JConsol

[Vadim Chekan <ko...@gmail.com>]

>From my experience, if you write a test case and create a JIRA issue,
then there are good chances it'll be fixed.

Vadim.

On Thu, Nov 13, 2008 at 11:55 PM, WhiteSock <wh...@hotmail.com> wrote:
>
> When I use JConsole to monitor ActiveMQ5.1, I find that the
>
> MessageCountAwaitingAcknowledge and DispatchedQueueSize of SubscriptionView
> are negative
>
> from time to time, especially when the producer is very fast and the
> consumer is relatively slow.
>
> This is the screen shot
> http://www.nabble.com/file/p20496111/jconsole.jpg
>
> Accoring to my test case, I think the EnqueueCounter is right. The
> DequeueCounter does not
>
> look right as well. I appologize if this issue has been raised before.
> --
> View this message in context: http://www.nabble.com/Negtive-MessageCountAwaitingAcknowledge---DispatchedQueueSize-on-JConsole-tp20496111p20496111.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>
>



-- 
>From RFC 2631: In ASN.1, EXPLICIT tagging is implicit unless IMPLICIT
is explicitly specified

Re: Negtive MessageCountAwaitingAcknowledge & DispatchedQueueSize on JConsol

[WhiteSock <wh...@hotmail.com>]

This problem still exists in ActiveMQ 5.2 
-- 
View this message in context: http://www.nabble.com/Negtive-MessageCountAwaitingAcknowledge---DispatchedQueueSize-on-JConsole-tp20496111p20679908.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

Re: Negtive MessageCountAwaitingAcknowledge & DispatchedQueueSize on JConsol

[WhiteSock <wh...@hotmail.com>]

I was able to reproduce this problem on ActiveMQ 5.1

1. Run SimpleBroker
   Please run MySql first.(I want to use JDBC Master/Slave cluster in
production environment)

2. Run SimpleConsumer
   In SimpleConsumer.java, it actually starts two MessageConsumers:
   * the first MessageConsumer is blocked in onMessage() by setting
consumers[i].setTimeToSleep(Integer.MAX_VALUE);
   * the other one won't be blocked by setting
consumers[i].setTimeToSleep(0);

   These consumers use the following broker url:
  
"failover:(tcp://localhost:61616,tcp://localhost:61617)?initialReconnectDelay=100&randomize=false"; 
   tcp://localhost:61616 is connectable and tcp://localhost:61617 is not.

3. Run SimpleProducer
   Send text message in a very fast speed.

4. After several minutes, the following exceptions were printed in
SimpleConsumer's Log:
Exception in thread "ActiveMQ Transport: tcp://localhost/127.0.0.1:61616"
java.lang.OutOfMemoryError: Java heap space
Exception in thread "ActiveMQ Transport: tcp://localhost/127.0.0.1:61616"
java.lang.OutOfMemoryError: Java heap space

   The screenshot at that moment is:
     http://www.nabble.com/file/p20679566/5.1-A.jpg 

   Several seconds later,  MessageConsumers seemed to be able to reconnect
to the broker.
   The log of SimpleConsumer is:
   2008-11-25 19:13:06,750 WARN  [InactivityMonitor Async Task:
java.util.concurrent.ThreadPoolExecutor$Worker@1d7ad1c -
FailoverTransport.java:205] Transport failed to tcp://localhost:61616 ,
attempting to automatically reconnect due to:
org.apache.activemq.transport.InactivityIOException: Channel was inactive
for too long: localhost/127.0.0.1:61616
2008-11-25 19:13:06,859 WARN  [InactivityMonitor Async Task:
java.util.concurrent.ThreadPoolExecutor$Worker@1d7ad1c -
FailoverTransport.java:205] Transport failed to tcp://localhost:61616 ,
attempting to automatically reconnect due to:
org.apache.activemq.transport.InactivityIOException: Channel was inactive
for too long: localhost/127.0.0.1:61616
[Consumer1]transport resumed
2008-11-25 19:13:08,140 INFO  [ActiveMQ Task - FailoverTransport.java:693]
Successfully reconnected to tcp://localhost:61616
[Consumer0]transport resumed
2008-11-25 19:13:08,140 INFO  [ActiveMQ Task - FailoverTransport.java:693]
Successfully reconnected to tcp://localhost:61616
       The screenshot after reconnection is:


-- 
View this message in context: http://www.nabble.com/Negtive-MessageCountAwaitingAcknowledge---DispatchedQueueSize-on-JConsole-tp20496111p20679566.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

Re: Negtive MessageCountAwaitingAcknowledge & DispatchedQueueSize on JConsol

[WhiteSock <wh...@hotmail.com>]

Thank you very much for your advice.

I have tried ActiveMQ5.2 RC2, It seems that the problem still exists.

I will raise a test case ASAP.


vchekan wrote:
> 
> From my experience, if you write a test case and create a JIRA issue,
> then there are good chances it'll be fixed.
> 
> Vadim.
> 
> On Thu, Nov 13, 2008 at 11:55 PM, WhiteSock <wh...@hotmail.com>
> wrote:
>>
>> When I use JConsole to monitor ActiveMQ5.1, I find that the
>>
>> MessageCountAwaitingAcknowledge and DispatchedQueueSize of
>> SubscriptionView
>> are negative
>>
>> from time to time, especially when the producer is very fast and the
>> consumer is relatively slow.
>>
>> This is the screen shot
>> http://www.nabble.com/file/p20496111/jconsole.jpg
>>
>> Accoring to my test case, I think the EnqueueCounter is right. The
>> DequeueCounter does not
>>
>> look right as well. I appologize if this issue has been raised before.
>> --
>> View this message in context:
>> http://www.nabble.com/Negtive-MessageCountAwaitingAcknowledge---DispatchedQueueSize-on-JConsole-tp20496111p20496111.html
>> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>>
>>
> 
> 
> 
> -- 
> From RFC 2631: In ASN.1, EXPLICIT tagging is implicit unless IMPLICIT
> is explicitly specified
> 
> 

-- 
View this message in context: http://www.nabble.com/Negtive-MessageCountAwaitingAcknowledge---DispatchedQueueSize-on-JConsole-tp20496111p20613361.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.