Hash address Generation

George Penokie gop at us.ibm.com
Mon Oct 9 11:19:02 PDT 2006


Formatted message: <A HREF="r0610093_f.htm">HTML-formatted message</A>

Sandeep,
We know that the hash addresses will duplicate. It is not possible for 
every 64 SAS address value to be uniquely identified in a 23 bit value. 
It's a question of probabilities. 
So given that the SAS addresses have to follow a certain structure to be 
valid there is a very low probability that two SAS addresses in the same 
system will hash to the same value and, even if that did happen, that a 
frame between those two would get misdirected.
By the way your example cannot happen with a valid SAS address.
Bye for now,
George Penokie
Dept 9A8 030-3 A410
E-Mail:    gop at us.ibm.com
Internal:  553-5208
External: 507-253-5208
Sandeep taneja <sandeep.taneja at nsysinc.com> 
Sent by: owner-t10 at t10.org
10/09/2006 11:53 AM
To
t10 at t10.org
cc
Monika <monika.Talwar at nsysinc.com>
Subject
Hash address Generation
* From the T10 Reflector (t10 at t10.org), posted by:
* Sandeep taneja <sandeep.taneja at nsysinc.com>
*
Hi all
I have a little doubt regarding the Hash Address generation Logic
Ques:  The following two 64 bit address maps to same 24 bit address
64 bit address value   23 bit hash adress 
00000000_00000001h	DB2777h
FFFFFFFF_FFFFFFFFh	DB2777h
Then How Hash address is used to check that whether frame is routed
properly or not , as from this we can assume that Hash Generation Logic
is not One to One (64 bit address maps to unique 24 bit hash address
value).
Plz suggest.........
regards 
Sandeep
*
* 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