kahadb lock in logs

[andrewhelicopter <an...@petcircle.com.au>]

This has been occuring since version 5.10 for us. It appeared that if we run
activemq as a different user than who owns the activemq files then the
kahadb will lock up and activemq will crash.

So we've made sure we chown user:user -R ./activemq
then run (without sudo)
./activemq/bin/linux-x86-64/activemq start

That's stopped activemq from crashing out but we still are seeing this every
4 hours like clockwork.

2015-02-19 10:48:57,686 | INFO  |
PListStore:[/opt/apache-activemq-5.11.0/data/localhost/tmp_storage] started
| org.apache.activemq.store.kahadb.plist.PListStoreImpl | main
2015-02-19 10:48:57,723 | INFO  | Using Persistence Adapter:
KahaDBPersistenceAdapter[/opt/apache-activemq-5.11.0/data/kahadb] |
org.apache.activemq.broker.BrokerService | main
2015-02-19 10:48:57,735 | INFO  | Database
/opt/apache-activemq-5.11.0/data/kahadb/lock is locked... waiting 10 seconds
for the database to be unlocked. Reason: java.io.IOException: File
'/opt/apache-activemq-5.11.0/data/kahadb/lock' could not be locked. |
org.apache.activemq.store.SharedFileLocker | main




--
View this message in context: http://activemq.2283324.n4.nabble.com/kahadb-lock-in-logs-tp4691746.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

Re: kahadb lock in logs

[andrewhelicopter <an...@petcircle.com.au>]

Replying to my own post now months later.

We successfully had 5.11.0 running for over 136 days until we decided to do
some maintenance and reboot the server this morning.

Now first problem, is again we couldn't get the lock. The consumers would
connect but could not enqueue messages. I will attach the log file.
activemq.log-old
<http://activemq.2283324.n4.nabble.com/file/n4698488/activemq.log-old>  

So I moved the old directory then extracted a fresh copy from
apache-activemq-5.11.0-bin.tar.gz and tried to start. Again there was a
problem attaining lock. Deleted the lock file and the suddenly the
connections could send messages !

The problem I'm seeing now is that the broker tries to start another service
but cannot due to the existing process.
activemq.log-new
<http://activemq.2283324.n4.nabble.com/file/n4698488/activemq.log-new>  

Is this a bug or what ?



--
View this message in context: http://activemq.2283324.n4.nabble.com/kahadb-lock-in-logs-tp4691746p4698488.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.