RE: iSCSI: INQUIRY page 0x83 identifier

["Elliott, Robert" <Robert.Elliott@compaq.com>]

> -----Original Message-----
> From: Michael Krause [mailto:krause@cup.hp.com]
> Sent: Friday, January 26, 2001 9:08 PM
> To: Robert Snively
> Cc: ips@ece.cmu.edu
> Subject: RE: iSCSI: INQUIRY page 0x83 identifier
> 
...
> Why not just require EUI-64 which is a very large name space 
> and is / will be used in many hardware solutions today / future?
> This would also fit in well with other I/O standards that use
> EUI-64 for identifying all I/O components / controllers.  Having
> a 128-bit value seems like overkill and provides little benefit.

A RAID controller needs to construct unique identifiers for
its virtual volumes.  The hardware typically has a single
EUI-64.  It's difficult for software to construct unique
64-bit identifiers that don't interfere with the EUI-64
namespace.  The hardware EUI-64 would have to represent a
range rather than just one ID.  It's much simpler to use the 
EUI-64 as a base and append bits to the end.  That's how the 
128-bit Registered Extended format is used.

I wouldn't mind only documenting the 64-bit IEEE Registered
and 128-bit IEEE Registered Extended formats in type 3h.
However, this is an SPC-3 issue, not an iSCSI issue.

The difference between the type 2h EUI-64 and the type 3h
IEEE Registered type is the latter includes a 4-bit Naming
Address Authority (NAA) field, restricting the 
vendor-specific portion to 36 bits.

Type 2 (EIU-64):
	24-bit company ID + 40 bit vendor-specific

Type 3 (FC-FS): 
	4 bit NAA + 60 bit NAA-specific

  Type 3 NAA=IEEE Registered:
	24-bit company ID + 36-bit vendor-specific
  Type 3 NAA=IEEE Registered Extended: 
	24-bit company ID + 36-bit vendor-specific + 64-bit vendor-specific

---
PC: Robert.Elliott@compaq.com
UNIX: relliott@unixmail.compaq.com