FC-S-2 to obsolete NAA type 4

Matthew Jacob mj at feral.com
Mon Oct 11 19:10:02 PDT 2004


* From the T10 Reflector (t10 at t10.org), posted by:
* Matthew Jacob <mj at feral.com>
*

Hmm. How amusing. I was at one using type 4 specifically in cases that I 
knew the primary IP of a system but had scrogged NVRAM for older QLogic 
2100 cards. This is an edge case.

It is wortwhile to note that Type 4 NAA gives you something from another 
name domain which is at least slightly unique with respect to the bundle 
of machines you are likely to be working with. Otherwise you have to try 
and derive a 64 bit unique quantity from a plethora of hardware 
(processor ID register; 48 bit mac address) or software (UUID 
generators).

One could argue that the 48 bit mac address of a primary ethernet 
interface is certainly more unique than an IP address. I would only note 
that it's quite possible to have a relatively unique and usable IP 
address without any ethernet interfaces.

All that said, I wouldn't be as upset about this change as others that 
have been proposed recently.

On Mon, 11 Oct 2004 Bob.Nixon at emulex.com wrote:

> Those present at the FC-FS-2 ad hoc meeting 5 October 2004 (28 people representing 26 companies), having found no evidence of use in existing or past products of NAA type 4 Address_Identifiers (i.e., those based on IP version 4 addresses), unanimously recommended obsoleting NAA type 4.
>
> Supported response to this message may be grounds for reconsideration ot that action.
>
>   - Bob Nixon, FC-FS-2 secretary
>
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org




More information about the T10 mailing list