activemq-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From goldi <>
Subject Apache ServiceMix – ActiveMQ Crash!
Date Wed, 14 Nov 2007 09:36:19 GMT

Hy everybody,

we are using Apache - Servicemix. In our Test-Environment (WinXP, CPU 2,8
GHz, 2,0 GB RAM) everything works fine, also under heavy throughput. Running
nearly the same Configuration in our productive environment: Windows Server
2003 x64 Edition, CPU 3,0 GHZ, 8,00 GB, Java “jdk1.5.0_13_64” RAM the
ESB/ActiveMQ crashes under the same throughput as in the Test-Environment.
One change we made in the productive environment is in the service.bat:

REM set DEFAULT_JAVA_OPTS=-server -Xmx512M
REM  Minimun 6,5GB Max 7GB, 512MB MaxPermSize 
set DEFAULT_JAVA_OPTS=-server -Xms6656m -Xmx7168m -XX:MaxPermSize=512m

The question is what makes the ESB/ActiveMQ crash? Could it be the
environment, could be the change in servicemix.bat., could be the x64
java-version? I made many test under heavy load in the Test-Environment
before and never got the ESB/ActiveMQ to crash.
Another heavy problem which occurred after this crash: We started the ESB
again and all of our (Queue) Consumers (JBoss with ActivemMQ-Broker) were
reconnected, but some of them stopped/don’t started consuming messages.
After restarting the Consumers they started consuming again. We were able to
reproduce this problem in our Test-Environment.  We just killed the ESB
process and then started the ESB again. Some of the consumers were able to
consume message some of them not. So in the case of another crash of the ESB
we would have to restart all consumers, but this is not acceptable.

We are using Servicemix 3.1.2. with ActiveMQ 4.1.1 (with default persistence
Adapter journaledJDBC).
Two error-files of the JVM are appended. 

Greets goldi

PS: Post this to ActiveMQ and to Servicemix. hs_err_pid3372.log hs_err_pid4492.log 
View this message in context:
Sent from the ActiveMQ - User mailing list archive at

View raw message