Re: iSCSI : versions and draft no.s (fwd)

["Robert D. Russell" <rdr@io.iol.unh.edu>]

Julian, Subrahmanya:

The consensus amoung consortium members for the next plugfest
was NOT to implement an X- key for this extension, but rather
to agree to configure the implementations for a specific
draft prior to testing and then have login requests and
responses carry in the their version-max,
version-min, version-active fields only values allowed by
that draft (so for drafts 11 and after those fields could
contain only 0).

This means that targets and initiators can not adapt
dynamically to the other side of the connection,
(they are statically configured to 1 draft only) but for
the plugfest environment this was not considered essential.
The overriding consideration was that implementing an X-
key was a waste of everybody's time, since it would
become invalid as soon as the standard was approved.


Bob Russell
InterOperability Lab
University of New Hampshire
rdr@iol.unh.edu
603-862-3774

On Tue, 21 May 2002, Julian Satran wrote:

> That is still the case - Julo
>
>
>
>
> Subrahmanya Sastry K V <skotra@npd.hcltech.com>
> Sent by: owner-ips@ece.cmu.edu
> 05/21/2002 03:15 PM
> Please respond to Subrahmanya Sastry K V
>
>
>         To:     ips@ece.cmu.edu
>         cc:
>         Subject:        iSCSI : versions and draft no.s
>
>
>
> Hi all,
>
> I would like to know the final word on the earlier discussion on the
> absence of a clear mapping between the iscsi draft version no.s and the
> various drafts.
>
> The last suggestion I saw in an earlier thread was that we could use a
> vendor specific X-draft=<draft version(s)>  key, values to indicate the
> supported draft versions. The same was seconded by Julo for the
> plugfest. Is this the final conclusion for an implementation to support
> multiple versions ?
>
> Thanks
> --
> Subrahmanya Sastry K V
> Member Technical Staff
> HCL Tech, Chennai, INDIA
> http://san.hcltech.com
>
>
>
>
>