On Wed, 2010-08-04 at 11:14 -0500, Cory wrote: > Looking over the list archives and Wiki there doesn't seem to be any > sort of built-in host redundancy with the proxy. I would prefer to use > a built in though it doesn't look like this is possible. > > One of the suggestions that I found in the list archives was to write a > script which checks the health of the hosts and adjusts the MySQL table > accordingly. Has anyone done anything like this in production?
Are you using NFS or some other shared filesystem? Then you should probably use the new director service in v2.0: http://blog.dovecot.org/2010/05/new-director-service-in-v20-for-nfs.html It doesn't support automatic failover yet either, but it's much easier to implement to it.
