Its been a while since the below post. So, i have a Qn in regards to Master-Slave clustering:
Is fail-back still an issue in Master-Slave clustering? Any idea when its gonna be fixed? Clustering using Shared (FS/DB) seems to be the only solution at this moment and is not a great idea for implementation for cross network or multi-data center implementation. I am not sure, but there's something surely missing somewhere: 1. I bought the ActiveMQ in Action E-book yesterday (8/27/2009) and the chapter related to CLUSTERING is missing or rather not yet WRITTEN!! (why?) 2. Since FUSE is also based on AMQ, i downloaded their doc.zip file and even there there is no reference to CLUSTERING. Also, the docs shows "Admin guide" in the index but the actual chapter is MISSING!! If anyone knows, please share with us. We cannot even think of implementing AMQ (or FUSE) in our org if it does not support proper industry standard clustering mechanism. -vezee ---------------------------------- For clustering, you use Master/Slave... http://activemq.apache.org/masterslave.html (not store and forward networks). You're main choice is between using a shared file system or database - for easy fail-back and management - or to use pure master slave which only supports failover, not fail-back and you have to manually copy around files if you want to bring back an old master. -- James ------- -- View this message in context: http://www.nabble.com/Broker-Clustering-tp19271212p25181443.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
