On 05/23/2012 12:05 AM, ryanL wrote:
on a similar note, how do people address the situation of a server
doing bond0 to two different top of rack switches, and a switch uplink
fails? in this situation, the two tor switches are not connected (i
dislike spanning tree). the bond0 interface can't see that uplink
failure,
There's your problem. Use a bonding method that can detect uplink failure.
I like the active-standby w/ ARP probes of the gateway, on Linux (which
given you say "bond0" is, I assume, what you mean)
As others have pointed out, if you can't do this then some triggered
event, possibly via EEM, might work. But frankly I dislike those
solutions - "smarts in the network" versus "smarts at the edge" and all...
_______________________________________________
cisco-nsp mailing list cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
