On 5/20/09 3:54 PM, "Tom Duerbusch" <[email protected]> wrote:
> When mainframes just started getting into IP communications, one of the things > we were warned about is to filter out any unnecessary IP traffic that is not > for the same subnet that the mainframe was on. The reason that I remember, is > that the IP stack looked at every packet to see if that packet was destined > for itself. Hence an interrupt was generated for each packet sent on the LAN. Had more to do with the fact that the Ungerman-Bass ethernet cards in the older devices would just fall over and stop running if there was too much junk on the wire. The 8232 and the 7170 were particularly sensitive to that. The 3172 was better behaved, but could still be overwhelmed if you got a good ARP storm going. So, best practice was to isolate them onto a separate segment so the only thing that could screw them up. > Ok, now we are on VSWITCH and Guest Lans. Does that same problem exist within > the VSWITCH/Guest LANS? Possibly on layer 2 VSWITCH, where the guest stack has to wake up and do some of the discrimination. Layer 3 GLANs and VSWITCHes shouldn't suffer from this.
