I've read some scarey technical questions and answers before but this is terrifying; it's a rather blase answer to a question about merging SAN fabrics with switches from different vendors. Folks, if you can possibly avoid this, avoid it!
It is rarely a good idea and if you are using any of the functionality that you have probably paid for, it is going to be a horrible job! You would better implementing a seperate fabric in most cases as opposed to growing an existing fabric with a different vendor's switches.
Believe me, I've looked at in the past and the certification matrices, patches and the added complexity means that any short term saving disappears very quickly! Actually, the lack of interoperability between the various FC vendor's kit is a reason to go for iSCSI for many people.
I’d agree that that is one of the most glib Q&As on one of the most complex tasks I’ve seen in many a year!
But … cross vendor merges are possible (and very necessary for some customers in these times of consolidation and merger), but should be one of the most researched, planned and carefully executed operations in the Data Center runbook. In a perfect world I’d lean toward your solution of a migration of resources to a new fabric, but, the world is rarely that perfect.
We’ve successfully merged fabrics for customers on numerous occassions, but its a job not attempted lightly and the prep vs implementation/test/handover is immensely skewed into the former stages. I’ve been successfully delivering a taste of these issues in training classes also for many years including McData, QLogic, Brocade and Cisco cross-vendor fabrics (the joys of being an independent and a mixed customer base!), to make people consider the issues in a benign environment.
In short, its not somewhere your grand plan would have have you be, but its a situation that with planning can be overcome, enabling you to chart a better course to calmer seas.
Craig
solid gold….