It was committed and works, but the point is that using it as described in the Jira commit (i.e. using it in a HA failover cluster) has strange behaviour. As Anthony described, it's actually best not using this and using separate hostnames for each half of the cluster, the failover still works perfectly as both switches read pertinent data from a shared database even though their hostnames are different. There may be other reasons for using a switchname which is different to the machine's hostname and if so, then switchname in switch.conf.xml works as described. I've gone back to my test setup and am re-testing failover in test conditions, when I've put this into a live scenario i.e. everything works I'll look at the Wiki article and see if anything needs updating, although Avi has rightly noted there that the failover works with different hostnames. Neil. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/HA-Starting-FreeSWITCH-empties... Sent from the freeswitch-users mailing list archive at Nabble.com.