RE: storage-device QoS [was: IPS Draft Charter update]

["Howard Hall" <hhall@ultranet.com>]

The comment should have read "...will not cut it ALONE in the more
complicated.." and was in reference to Mike's comments about just mapping of
priority reqs to these fields.  I've read 2474 in the past, and I admit I am
rusty on it, I will read it again. From what I remember, some of the more
complicated concepts of QoS we have thought of here at Pirus will need to
extend into the protocol, in some cases complementing the DS fields.

Dave: How widespread is the adoption and use of 2474?

-Howard

Howard Hall
Pirus Networks
www.pirus.com

-----Original Message-----
From: Brian E Carpenter [mailto:brian@hursley.ibm.com]
Sent: Monday, July 24, 2000 10:21 AM
To: hhall@ultranet.com
Cc: 'Michael Krause'; ips@ece.cmu.edu
Subject: Re: storage-device QoS [was: IPS Draft Charter update]


Howard Hall wrote:
...
> It is also
> clear the TOS or TClass fields will not cut it in the more complicated
(and
> desirable) QoS concepts.

This is a pretty unclear comment. Apart from anything else, the IETF
obsoleted
the old concept of the TOS field in RFC 2474. The replacement concept of the
Differentiated Services field (which applies identically to IPv4 and IPv6)
is considerably more complex and subtle, and goes way beyond the notion
of simple priority that has been mentioned on this thread.

I agree that this question should be deferred to a later stage.

  Brian