Mailing-List: contact users-help@activemq.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@activemq.apache.org
Received-SPF: pass (nike.apache.org: domain of lists@nabble.com designates
 216.139.236.158 as permitted sender)
Message-ID: <25094579.post@talk.nabble.com>
Date: Sat, 22 Aug 2009 08:01:42 -0700 (PDT)
From: suresh_123456 <sanantpurkar@yahoo.com>
To: users@activemq.apache.org
Subject: ActiveMQ5.2 issues
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit


Hi - We are using ActiveMQ 5.2 in our production environment, recently
upgraded from 4.1 to 5.2 - here are some of the issues we ran into in 4.1
and current set of issues in 5.2

Ours is a high volume txn processing platform - activemq is used to store
and forward the messages to different applications - ontime delivery of
messages and zero loss of messages is critical to the success of our
business, any message that does not reach the destination with in 2-3 mins -
will result in failed txn for the customer who is sending the message ...


Issues faced in  4.1
                
1.	Registered listener loses the connection if there is no message on the
listening queue name. This is resolved by setting out a time out. If there
is no message on the queue for some stipulated time the listener reconnects
to Activemq
- do not know if this is the right setup 

2.	Setting up  maxInactivityTimeOUt=0 in the activemq.xml in Activemq
server, this resolved almost all problems except that it is not closing
connection at the server end. Hence after few days the connections used to
increase beyond threshold, hence we were bound to restart the server. The
most weird point is the no of connections in client (in DMZ zone) were
always in control (say constantly at 20)but if we check the no of
connections from the server end it would be high (say 100 and increasing). 
- i believe this is fixed in 5.2 - pl confirm

3.	Removing maxInactivityTimeOUt=0 in the URL maintained the connection
constant , but on load the connections suddenly used to go up as high as
1000.


In 5.2

1.	Duplicate message were created in the queue, this was resolved after
applying a patch (ACTIVEMQ)
2.	Initially we used AMP format for persisting the data, even it created out
of memory problem. Currently we are using Journal/JDBC for persistence 
- we felt AMP gave us better performance - but not had to revert to journal
/ jdbc - need to know which is the best way to get maximum throughput /
better response time.
3.	Like 4.1 again connections are increasing in the above similar way if we
set maxInactivityTimeOUt=0, but it is reducing . Before concluding, we need
to still keep a watch on the same if it reduces or not. 

Other questions - can we use clustering - we have application in DMZ and
internal zone and both connect to ActiveMQ

Hope to hear from you...

-- 
View this message in context: http://www.nabble.com/ActiveMQ5.2-issues-tp25094579p25094579.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.