On 16 Apr 2008, at 20:18, Bruce Hurd wrote:
>
> Thanks for the reply Rob,
> We haven't tested the java client but heres some further updates to
> the
> problem:
>
> testing against 5.1 RC3, sometimes there is an exception from the
> ActiveMQScheduler thread (see below) which i assume is the file
> cleanup
> process. When I saw this and restarted the broker, it reported no
> data
> files were in use and deleted them all. In further executions, this
> exception didnt occur and the files are being cleaned up as
> expected. So,
> using 5.1 RC3 and the latest build of the NMS project, expected
> behaviour is
> occurring sometimes.
>
> Any idea why this exception would be thrown, or what would be the
> best way
> to monitor for exceptions from the ActiveMQ Scheduler and restart
> the broker
> when they occur?
>
> Cheers,
> Bruce.
>
> Exception in thread "ActiveMQ Scheduler"
> java.util.ConcurrentModificationException
> at java.util.HashMap$HashIterator.nextEntry(HashMap.java:787)
> at java.util.HashMap$KeyIterator.next(HashMap.java:823)
> at java.util.AbstractSet.removeAll(AbstractSet.java:143)
> at
> org
> .apache
> .activemq
> .kaha
> .impl
> .async
> .AsyncDataManager.consolidateDataFilesNotIn(AsyncDataManager.java:404)
> at
> org
> .apache
> .activemq
> .store.amq.AMQPersistenceAdapter.cleanup(AMQPersistenceAdapter.java:
> 417)
> at
> org.apache.activemq.store.amq.AMQPersistenceAdapter
> $3.run(AMQPersistenceAdapter.java:269)
> at
> org
> .apache
> .activemq.thread.SchedulerTimerTask.run(SchedulerTimerTask.java:33)
> at java.util.TimerThread.mainLoop(Timer.java:512)
> at java.util.TimerThread.run(Timer.java:462)
>
>
>
> rajdavies wrote:
>>
>>
>> On 16 Apr 2008, at 19:17, Bruce Hurd wrote:
>>
>>>
>>> Does anyone have any ideas why the data files in the journal folder
>>> are never
>>> deleted.
>>>
>>> Test:
>>>
>>> Install ActiveMQ 5 / 5.1 on Windows platform
>>> Modify default config not discover / network with other brokers, and
>>> set the
>>> data file size of the amqPersistenceAdapter to 1mb (so we get lots
>>> of files)
>>>
>>> publish 10000 messages at about 500K each using the ActiveMQ .net
>>> client to
>>> a queue.
>>> Subscribe to the queue using the .net client with ClientAcknowledge
>>> and
>>> attach a listener.
>>> Receive all the messages, acknowledging each one and close client.
>>> Check the queues attributes on JConsole (enqueue 10000, dequeue
>>> 10000, queue
>>> size 0)
>>>
>>> wait for the data files to be deleted. This never happens!!
>>>
>>> Any idea? This is making us run the broker in non persistent mode
>>> (not
>>> good!) till we find a solution.
>>>
>>> Thanks in advance,
>>>
>>> Bruce.
>>>
>>>
>>> --
>>> View this message in context:
>>>
http://www.nabble.com/AMQ-Journal-data-files-never-deleted-tp16728103s2354p16728103.html>>> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>>>
>> Hi Bruce,
>>
>> do you get the same behavior with a java client ? Just trying to
>> narrow the domain down a little - as I haven't seen the same
>>
>>
>>
>>
>> cheers,
>>
>> Rob
>>
>>
http://open.iona.com/ -Enterprise Open Integration
>>
http://rajdavies.blogspot.com/>>
>>
>>
>>
>>
>
> --
> View this message in context:
http://www.nabble.com/AMQ-Journal-data-files-never-deleted-tp16728103s2354p16731446.html> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>
I raised an issue for this:
https://issues.apache.org/activemq/browse/AMQ-1669hopefully the fix will make tomorrow's snapshot release for 5.2
cheers,
Rob
http://open.iona.com/ -Enterprise Open Integration
http://rajdavies.blogspot.com/