I would appreciate some expert advice on setting up my ESX servers to a fibre channel frabic.
The SAN is a IBM DS3400 Active/Passive array. We have about 6 LUNS. The FC switches are Cisco MDS 9000 series.
Attached is a diagram of how I think it should be configured based on best practices. Please click on the thumbnail.
As you can see there are two separate fabrics and two zones on each switch, the reason for that is I believe that the HBA’s shouldn’t be on the same zone and neither should the two storage processors.
For example an ESX server can see a LUN via the following 4 paths:
SP-A via HBA1, fabric 1 / zone 1
SP-A via HBA2, fabric 2 / zone 3
SP-B via HBA1, fabric 1 / zone 2
SP-B via HBA2, fabric 2 / zone 4
Now let’s say LUN1 is currently owned by SP-A, an ESX server can use the following HBA’s:
SP-A via HBA1, fabric 1 / zone 1
SP-A via HBA2, fabric 2 / zone 3
If HBA1 was to fail it could start using HBA2 without causing the Storage processors to flip
But if for example a Storage processor was to fail the ESX server could continue to use the same HBA.
Now this is the topology I have come up with based on the information I have found, but I would appreciate if anyone could confirm with me that this looks OK and if there is anything I should watch out for.
Nicholas