1
votes

I am attempting to run jboss AS 7 as a service on raspbian. Sometimes it starts without issue (20% of the time?) other times "Brontes" failed to start. The errors are not always the same. I have seen over half a dozen, usually regarding the loading of some jboss module.Here is an example:

Operation ("parallel-extension-add") failed - address: ([])
java.lang.RuntimeException: JBAS014670: Failed initializing module org.jboss.as.threads

It is not always the same module in the error. You can subsitute org.jboss.as.threads for org.jboss.as.messaging, org.jboss.as.security, org.jboss.as.logging, org.jboss.as.connector... etc. Another one that occasionally occurs is failed to parse configuration.

The failure occurs very early, before it ever gets to deployment of my .ear.

Here are the java_opts from my standalone.conf

JAVA_OPTS="-Xms64m -Xmx512m -XX:MaxPermSize=256m -Djava.net.preferIPv4Stack=true -Dorg.jboss.resolver.warning=true -Dsun.rmi.dgc.client.gcInterval=3600000 -Dsun.rmi.dgc.server.gcInterval=3600000 -Djboss.modules.system.pkgs=org.jboss.byteman -Djava.awt.headless=true -Djboss.server.default.config=standalone.xml"

When I start this directly by executing standalone.sh it seems to start consistently. The inconsistency appears to only happen when it is started as a service (esp. unattended at boot).

I thought that there was perhaps some race condition, but its too inconsistent and I know too little about JBoss and the environment I am in to diagnose.

The LSB header on my inint.d script:

### BEGIN INIT INFO
# Provides: jboss
# Required-Start: $local_fs $remote_fs $network $syslog
# Required-Stop: $local_fs $remote_fs $network $syslog
# Default-Start: 2 3 4 5
# Default-Stop: 0 1 6
# Short-Description: Management of JBoss AS v7.x
### END INIT INFO

I inherited this solution and I am not very familiar with JBoss at all.

1

1 Answers

1
votes

Closing this. No responses, no progress.
In the end I started with a fresh image for the Pi and reloaded. Running fine now. I don't want to jump to the "something got corrupted" explanation but that really seems to have been the case. It was so many different, unrepeatable errors under different scenarios and over time it started correctly less frequently.