We gave a lun to a ESX host and it just happened that because we used ECC that the LUN ID from each of the two FAs was not the same. In the prefect workd the lunids match but it is not realistic to expect that. Never had issue before on Physical server with that but the VM guys are complaining that tryng to add it a RDM it shows as two devices? is that an issue or a VMWARE thing. Hate to have to play with it to make the luns the same on both FAs if not needed.
i once had the same issue, i dont remember exactly what fixed it (it was long time back). I think i ran a device scan on ESX to fix it...i will check it again in the lab...
If your environment doesn't automatically match FA scsi IDs, you could always map them to the FA's with some high value like 3E8 and use symmask offset to show it to the host as SCSI ID 1.
if you are running at 5772 code or above you can specify a lun id to be used on both FA's. Using this method you can have the same ID for the 2 FA without an issue. One thing to Note is that if you start using this method you need to continue using it, you will be requested to use the -lun flag or dynamic_lun flag each time you wish to assing new devices, however once you know this is the case then it's not a problem.
There are a number of benefits to using the dynamic lun addressing as you are now using a soft lun id. This means you have the full address range 0-255 available for each FA and are not bound by the hard lun address on the FA. This is explained better than I can do it in the array controls guide.
Worth a look and hope it helps