Re: Out Of Memory Error while running TopicPublisher/TopicListener

[Rob Davies]

Your client is running foul of the topic prefetch - either use more  
memory (as you've suggested) - or use a smaller prefetch - see
http://activemq.apache.org/what-is-the-prefetch-limit-for.html.
The slow consumer advisory will be generated from the broker when this  
limit is reached - but only if enabled (its off by default)

To enable it - use a DestinationPolicy - see 
http://activemq.apache.org/per-destination-policies.html
The documentation here needs updating - but the property you need to  
set is advisoryForSlowConsumers

cheers,

Rob

Rob Davies
http://rajdavies.blogspot.com/


On 16 Sep 2008, at 22:01, Badri wrote:


I am running broker, publisher, consumer all in Windows.
I have not tried this in Linux.
The broker & publisher are working fine. It's the consumer which is
crashing. Removing Thread.sleep(1000), consumer works fine. It did  
not crash
for a good amount of time before I manually stopped the TopicListener.

Windows XP Professional with SP2, 2GB RAM.
Java: 1.6.0_06

Of course this was my first test to simulate slow consumer & also  
see if the
advisory topic appears in the web console. So I used my desktop &  
stumbled
across this memory error. I just felt maybe crash could be due to  
using
Windows. I intend to do the same test in Linux.


bsnyder wrote:

On Tue, Sep 16, 2008 at 10:33 AM, Badri
<[EMAIL PROTECTED]> wrote:

Hi

I am trying to simulate slow consumer using the classes
(TopicPublisher/TopicListener) provided in examples.

I wanted to see if Web Console prints the Advisory topic for Slow
Consumer.
But I am facing Out Of Memory Error & the consumer stops abruptly.

Following are changes I did:

activemq.xml:

      <transportConnectors>
          <transportConnector name="openwire"
uri="tcp://localhost:61616?maxInactivityDuration=-1"/>
      </transportConnectors>


TopicListener: I added a sleep of 1000 to induce some slowness.
          if (++count % 1000 == 0) {
              System.out.println("Received " + count + "  
messages.");
                              try
                              {
                                      Thread.sleep(1000);
                              }
                              catch (Exception e)
                              {

                              }
          }

I give java -Xmx256m -Xms128m when using TopicListener.
After maybe 5 minutes it crashes. After receiving anywhere between  
15-20
batches of 1000 messages, it crashes.

Exception:
Received 9000 messages.
Received 10000 messages.
javax.jms.JMSException: Unexpected error occured
      at
org 
.apache.activemq.util.JMSExceptionSupport.create(JMSExceptionSuppo
rt.java:62)
      at
org 
.apache.activemq.ActiveMQConnection.syncSendPacket(ActiveMQConnect
ion.java:1255)
      at
org.apache.activemq.ActiveMQSession.send(ActiveMQSession.java:1659)
      at
org 
.apache.activemq.ActiveMQMessageProducer.send(ActiveMQMessageProdu
cer.java:227)
      at
org 
.apache.activemq.ActiveMQMessageProducerSupport.send(ActiveMQMessa
geProducerSupport.java:241)
      at TopicListener.onMessage(TopicListener.java:96)
      at
org 
.apache.activemq.ActiveMQMessageConsumer.dispatch(ActiveMQMessageC
onsumer.java:983)
      at
org 
.apache.activemq.ActiveMQSessionExecutor.dispatch(ActiveMQSessionE
xecutor.java:122)
      at
org 
.apache.activemq.ActiveMQSessionExecutor.iterate(ActiveMQSessionEx
ecutor.java:192)
      at
org 
.apache.activemq.thread.PooledTaskRunner.runTask(PooledTaskRunner.
java:122)
      at
org.apache.activemq.thread.PooledTaskRunner 
$1.run(PooledTaskRunner.ja
va:43)
      at
java.util.concurrent.ThreadPoolExecutor 
$Worker.runTask(ThreadPoolExec
utor.java:885)
      at
java.util.concurrent.ThreadPoolExecutor 
$Worker.run(ThreadPoolExecutor
.java:907)
      at java.lang.Thread.run(Thread.java:619)
Caused by: java.io.IOException: Unexpected error occured
      at
org 
.apache.activemq.transport.tcp.TcpTransport.run(TcpTransport.java:
191)
      ... 1 more
Caused by: java.lang.OutOfMemoryError: Java heap space
      at
org 
.apache.activemq.openwire.v3.BaseDataStreamMarshaller.tightUnmarsh
alByteSequence(BaseDataStreamMarshaller.java:440)
      at
org 
.apache.activemq.openwire.v3.MessageMarshaller.tightUnmarshal(Mess
ageMarshaller.java:68)
      at
org 
.apache.activemq.openwire.v3.ActiveMQMessageMarshaller.tightUnmars
hal(ActiveMQMessageMarshaller.java:67)
      at
org 
.apache.activemq.openwire.v3.ActiveMQBytesMessageMarshaller.tightU
nmarshal(ActiveMQBytesMessageMarshaller.java:67)
      at
org 
.apache.activemq.openwire.OpenWireFormat.tightUnmarshalNestedObjec
t(OpenWireFormat.java:453)
      at
org 
.apache.activemq.openwire.v3.BaseDataStreamMarshaller.tightUnmarsa
lNestedObject(BaseDataStreamMarshaller.java:126)
      at
org 
.apache.activemq.openwire.v3.MessageDispatchMarshaller.tightUnmars
hal(MessageDispatchMarshaller.java:72)
      at
org 
.apache.activemq.openwire.OpenWireFormat.doUnmarshal(OpenWireForma
t.java:362)
      at
org 
.apache.activemq.openwire.OpenWireFormat.unmarshal(OpenWireFormat.
java:276)
      at
org 
.apache.activemq.transport.tcp.TcpTransport.readCommand(TcpTranspo
rt.java:209)
      at
org 
.apache.activemq.transport.tcp.TcpTransport.doRun(TcpTransport.jav
a:201)
      at
org 
.apache.activemq.transport.tcp.TcpTransport.run(TcpTransport.java:
184)
      ... 1 more
Received 1000 messages.
Received 2000 messages.

TopicPublisher:
private long batch(int msgCount) throws Exception {
// Commented out call to waitForCompletion
// waitForCompletion();
}

QN 1) I just want to know if I am configuring something wrong. Or  
should
I
increase -Xmx512m when running TopicListener to get rid of OOME?

You shouldn't need to raise the memory at all as there should be
nothing causing memory to grow. I just added the sleep above to the
TopicListener using ActiveMQ 5.2.0 and I've produced and consumed  
over
300,000 messages in a matter of moments without any issue whatsoever.
I'm on MacOS X using Java 1.5.0_13. What OS and Java version are you
using?

QN 2) Will the advisory topic:
ActiveMQ.Advisory.SlowConsumer.Topic.topictest.messages appear  
only if
the
consumer starts discarding messages? What is the basis for this  
topic to
appear in Web console?

I'm not able to see the *SlowConsumer* topic appear at all. I don't
know if that's because I'm using a dual core Intel processor or what.
What type of processor does your machine have?

Bruce
--
perl -e 'print
unpack("u30","D0G)[EMAIL PROTECTED]&5R\"F)R=6-E+G-N>61E<D\!G;6%I;\"YC;VT*"
);'

Apache ActiveMQ - http://activemq.org/
Apache Camel - http://activemq.org/camel/
Apache ServiceMix - http://servicemix.org/

Blog: http://bruceblog.org/



--
View this message in context: 
http://www.nabble.com/Out-Of-Memory-Error-while-running-TopicPublisher-TopicListener-in-ActiveMQ-5.2-tp19515522p19520251.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.