Cluster F/S: Network raid? OpenAFS? CODA? Fam / Imon triggered Rsync?
I cant really find a solution that does exactly what I want it to...
I have a load balanced / HA cluster (LVS/UltraMonkey) with 3 web server nodes initially with more to come. Each server is an exact replica of one another. Invoices are generated by out web based applications on any give application, but when it comes time to view the invoice, it needs to be available from any server you might hit.
Basically what im looking for is anything written to one of the cluster nodes to be replicated across to the other two. Id prefer if there were some sort of failback mechanism, but if it requires one node to be a server or manager, ill live with it.
What solution is best for me?
THANKS!!!
|