Greetings!
Just wanted to confirm my line of thinking on this with others. We want to
set up redundant radiator servers for our domain. We want to have a primary
and secondary, and NAS's will be told to check aaa in that order. If the
primary machine goes down, the secondary will still answer. We will be using
mySQL for the user database.
My thought was to have two machines, with each machine running both radiator
and mySQL. The radiator on the primary will use mySQL on the primary, the
radiator on the secondary will use mySQL on the secondary. This should
accomplish the above. Then we could set up radiator on the first machine to
use mySQL on the second machine (in addition) in case it's own mySQL process
fails and vice-versa on the secondary.
Several questions:
1) Is this a good recommended configuration or is there something I'm
missing or a better way to accomplish high availability? Do we need more
machines?
2) In the above config, the primary takes the full load and the secondary
only comes into play if the primary is down. In general terms, what changes
would need to be made to implement load balancing between the two instead
(with one machine taking the full load if the other fails)?
Thanks!
Jay West
===
Archive at http://www.thesite.com.au/~radiator/
To unsubscribe, email '[EMAIL PROTECTED]' with
'unsubscribe radiator' in the body of the message.
