>Hello everyone. > >Let me say up front, I'm no Cisco guru, although I do believe I posess a >sound understanding of networking involving multiple switches and the >potential issues associated with doing so. > >I'm looking at a situation where with the introduction of two machines >employing CARP to provide resiliant DNS services. The MASTER each would >hold an ip and either in the event of failure would hold both. It >worked fine for a little bit then all hell seemed to break loose on the >network. > >The network is 3 Catalyst 3750's "ringed" or "clustered" together. >There also exists on the network two 3.8 obsd pf's employing >carp/pfsync/ifstated. As well there exist several Linux boxes >performing LVS (VRRPv2 using same multicast address 224.0.0.18). > >For some reason, perhaps coincidence, when the CARP/DNS servers were >introduced great instability was observed until the CARP/DNS servers >were removed. > >Firstly is anyone aware of CARP + Cisco Catalyst switches 3750 or >otherwise involving single or multi carp scenarios (various pairs >performing different tasks on the same segment). [snip]
First, define the context of "great instability". Within the Cisco context? The Linux LVS context? The CARP context? Overall? Does this old thread makes sense? http://www.monkey.org/openbsd/archive/misc/0410/msg00867.html Anecdote: A few years ago, a large clustered Solaris environment I worked on started crashing when additional independent clusters were added. The cluster nodes talked via multicast. For a month, the network guys kept claiming "but the VLANs are private!" as an excuse and the server group retorted "but clearly the VLANs are not multicast-private!" When the mishandling of multicast on the 6513s was finally determined to be due to pilot error, the server team was able to stop using 2924s as cluster interconnects.
