.. what worries me at this point is introducing yet anothrer level of software complexity to solve what is really a simple requirement. I can see a simple database table with a little SQL code could solve this problem. What advantage does a JMS Queue (which is not a trivial software infrastructure to incorporate) give to this situation?
James.Strachan wrote: > > On 11/10/2007, Terry K <[EMAIL PROTECTED]> wrote: >> >> ... thanks for the suggestions - message groups do sound like a plausible >> solution. >> >> I am not interested in assigning a group of messages to particular >> sessions >> indefinately as the groups will be small and many. Once the high priority >> message has been dequeued with all related messages (at that point in >> time) >> that batch work is done and time to move onto the next high priority >> message. > > If you basically want to reorder a queue so that high priority > messages are processed first; with their related messages before > others, you maybe need this... > http://activemq.apache.org/camel/resequencer.html > > to reorder messages on the queue before they hit your consumer > > >> CorrelationID's also look promising. High do you create a message >> selector >> string to select by comparisons to other messages > > A JMS selector only performs a boolean expression on the current > message only - not on a window of previous messages > > >> e.g. CorrelationID and >> Priorty? I can see how this is done in the JMS Message javadoc API. All >> documentation seems to relate to literal comparisons > > See > http://activemq.apache.org/how-do-i-use-jms-efficiently.html > > with JMS if you want high performance you typically don't wanna be > starting/stopping subscriptions as you process individual messages. > > > -- > James > ------- > http://macstrac.blogspot.com/ > > Open Source SOA > http://open.iona.com > > -- View this message in context: http://www.nabble.com/message-selectors-tf4603950s2354.html#a13166731 Sent from the ActiveMQ - User mailing list archive at Nabble.com.
