broker discovery is completely separate to how a broker manages its
persistence; so any combination of transport & discovery should work
with all the master/slave implementations
On 4/25/07, Keith Irwin <[EMAIL PROTECTED]> wrote:
Hi folks--
We're using the 4.1.0 release....
I've been scanning the archives and docs and can't quite figure out if
I can use multicast with the shared filesystem master slave.
My aim is to NOT have to hardcode hostnames/ip addresses anywhere.
In my activemq config file, I have:
<persistenceAdapter>
<journaledJDBC dataDirectory="/our/shared/activemq"/>
</persistenceAdapter>
<transportConnectors>
<transportConnector name="openwire"
uri="tcp://localhost:61616"
discoveryUri="multicast://our-broker"/>
</transportConnectors>
And in my clients, I use:
"discovery:(multicast://keith)?initialReconnectDelay=100"
Will the shared filesystem master / slave technique work in this case?
(The docs mention:
failover:(tcp://broker1:61616,tcp://broker2:61616,tcp://broker3:61616)
but this requires us to hardcode our machine names/addresses which we
don't want to do).
Keith
--
James
-------
http://macstrac.blogspot.com/
