Hi All, I set up a testing environment to see if WSN Notification service is utilizing activemq's persistence capability, and it seems that it's not the case. It seems that when the consumer is down, the publishing of the message will fail.
In my environment, I have three parts: 1) Code to start WSN Notification service; 2) Code to publish a notification; 3) Code to subscribe to the notification as well as consuming the messages. I tried the following sequence: 1) Start the WSN Notification service; 2) Subscribe to the notification service with a topic; 3) Publish a message through the topic; 4) Shutdown the consumer of the message with specified topic; 5) Try to publish a message, but it failed. I got a connection refused exception. So, it seems that when publishing the message, WSN Notification service checks to see if the consumer is there or not. If it is not there, the publishing of the service will be failed. I was wondering why WSN Notification service is designed this way. Wouldn't it be better if the notification service can leverage on activemq's persistence capability to store the message when the flag is turned on, so that when the consumer is down, the message can be stored temporarily in activemq broker. When the consumer is up again, the message can be delivered to the consumer without losing it? Thanks very much, YuLing -- View this message in context: http://cxf.547215.n5.nabble.com/WSN-notification-service-does-not-utilize-activemq-s-persistence-capability-tp5613342p5613342.html Sent from the cxf-dev mailing list archive at Nabble.com.
