On 12/29/2010 02:49 PM, Kit Gerrits wrote: > Hello, > > AFAIK, Multi-interface heartbeat is something that was only recently > added to RHCS (earlier this year, if I recall correctly). > > Until then, the failover part was usually achieved by using a bonded > interface as heartbeat interface. > If possible, I would suggest using 2 (connected) Multicast switches and > running a bond from each server to each switch. > Or 2 regular switches and broadcast heartbeat (switches only connected > to eachother) > Otherwise, using an active-active bond (channel?) with 2 crossover > cables might also work, but offers less protection against interface > failures. > > > Regards, > > Kit Hi, It was around in el5. Perhaps not in the early versions, I am not sure exactly when it was added, but certainly by 5.4. In the recent 3.x branch, openais was replaced by corosync (for core cluster communications), which is where rrp is controlled. Of course, I could always be wrong. :) Cheers. -- Digimer E-Mail: digimer@xxxxxxxxxxx AN!Whitepapers: http://alteeve.com Node Assassin: http://nodeassassin.org -- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
