On 09/29/2010 08:35 PM, Paul Graydon wrote: > It's been a while since I've had to do anything much with networks, how > does stacking switches relate from an SPOF perspective? I would assume > that by reducing the number of cabinets / management points its > essentially a trade off for slightly less resiliency (lose one cab, lose > more of the building than you might otherwise have done so.) I presume > it's generally considered an acceptable gamble? > > Paul
The Cisco stacking architecture is a ring of up to 9 switches. If a single switch fails connectivity will be retained between the remaining switches, albeit at reduced bandwidth on the stacking bus. If the master switch in the stack fails, the rest of the stack will keep forwarding based on latest data while a new master is elected and takes over. (The stack master is responsible for all the management/control plane functions in the stack, so any of that function will be unavailable until a new master takes over.) Uplinks can also be spread across multiple switches (using STP, 802.3ad, etc.) so it's possible to maintain uplinks even in the event of a switch failure. Overall I don't see any significant loss of redundancy by stacking a number of switches in a rack vs having a collection of unstacked switches. (A double failure may result in a stack partition, but in my experience double failures have always been the result of environmental issues (water leaks and similar) that have ended up killing all/most of the stack and would have done so to unstacked switches also.) I don't know about stacking architectures from other vendors. -- Thanks Jefferson Cowart j...@cowart.net _______________________________________________ Tech mailing list Tech@lopsa.org http://lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/
