SAN design : core edge and dual-homing access switch

[Surya ARBY]

Hello all.

It may sound as a dumb question (from a LAN guy) but when designing a core/edge or edge/ecore/edge design, why do we connect access switches to both core switches ? Doesn't it break the isolation of a dual fabric backbone ?

If an access switch fails the fault (bug or anything else) will propagate to both core switches ? Am I wrong ?

Example :

http://www.cisco.com/en/US/prod/collateral/modules/ps5991/prod_white_paper0900aecd8044c807_ps5990_Products_White_Paper.html

or from netrworkers sessions in 2006

[krun_shah · Accepted solution]

Answer also from LAN guy,

Most likely this design diagram is due to assumption that there is no use of VSANs and SAN Multipathing drivers in host.

Following is excerpt from same like yo posted.

"SAN designs should always use two isolated fabrics  for high availability, with both hosts and storage connecting to both  fabrics. Multipathing software should be deployed on the hosts to manage  connectivity between the host and storage so that I/O uses both paths,  and there is non-disruptive failover between fabrics in the event of a  problem in one fabric. Fabric isolation can be achieved using either  VSANs, or dual physical switches. Both provide separation of fabric  services, although it could be argued that multiple physical fabrics  provide increased physical protection (e.g. protection against a  sprinkler head failing above a switch) and protection against equipment  failure. "

[Surya ARBY]

ok; so in the case of using standard dualport HBA (or CNA) the pictures above represent only one fabric:)

thank you !