On 5/9/06, Naby <[EMAIL PROTECTED]> wrote:
Hi. How must a network of brokers be configured? I found different examples for it. For example i have two brokers: A and B What is the difference between <networkConnector uri="static:(tcp://A:62626)" failover="true"/> and <networkConnector uri="static:(tcp://A:62626,tcp://B:62627)" failover="true"/> ?
Hi I'm not expert but I could share my experience and may be from activemq team could make a correction if I do a mistake. You need to configure in this way: <networkConnector uri="static:(tcp://A:outputport,tcp://B:transportport)"> failover="true"/> I mean, the first data is the output port of the broker you are configuring and the others are server to wich you one to connect and its transport ports (because brokers receive data from transport)
And must the two brokers both be written in the client conf like this: failover:tcp://A:62626,tcp://B:62627
You must list in client al servers you want to connect with its transport ports. In this way, client will try to connect randomly (as far as I remember) but you can force to assure to connect in the order you wrtie them.
If the brokers should be started on the same computer with different ports do i need run only one activemq.(sh or bat) ti start the brokers or must i run 2 brokers in different directories?
AFAIK, two brokers in different directories or at least with different journals. Also, if you are using persistence in DD, they should use different DB's. Hope this helps and sorry if I did a mistake with my explanation -- Javier Leyba Barcelona - Spain http://blog.leyba.com.ar
