I have no problems with Rajiv's proposed changes.  I think that
is better than revving 3291bis and affecting other mibs.

Brian

Dave Thaler wrote:

I also prefer it the way it is now, and Rajiv's proposed mods
look fine to me.

-Dave


-----Original Message-----
From: [EMAIL PROTECTED]

[mailto:[EMAIL PROTECTED]

bs.de] On Behalf Of Juergen Schoenwaelder
Sent: Monday, August 23, 2004 6:37 AM
To: Kristine Adamson
Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]
Subject: Re: [ipv6mib] Re: Auth48 update to TCP-MIB(draft-ietf-ipv6-
rfc2012-update-06.txt)

On Mon, Aug 23, 2004 at 04:04:57AM -0600, Kristine Adamson wrote:


In this case, would it be better to update draft-3291bis to permit
zero-length octet-strings for InetAddress objects whose

InetAddressType

is

IPv4 or IPv6?  We have already encountered problems with network
management applications not being able to get as much SNMP data back

in

an

SNMP response as they would like to be able to. With an IPv6

address in

an instance already taking up a lot of space in the SNMP response

packet,

it doesn't seem like a good use of response packet space to return

16

bytes of zeros for the address.

While I do not really buy into the bandwidth argument, I am indeed wondering whether the current rule is perhaps overly restrictive. At the moment (and that is unchanged from RFC 3291 and even from RFC

2851),

we have the following implication:

        InetAddress zero-length  =>  InetAddressType == unknown

To change this (basically removing this rule), we would have to remove
text from the 'unknown' definition and we would have to add text to

the

InetAddress TC that zero-length values can be legal subject to the
semantics spelled out in the definition where the TC is being used
(with a statement that per default they are not legal if nothing is
spelled out in the description clause (backwards compatibility)).

I am not sure which impact such a change will have. These TCs have
seen quite some usage - not only in the IETF, but also in vendor
specific MIB modules. And this is the first time I can remember
(and that I could find in my mailing list archive) that a change
to generally allow zero-length InetAddress values has been proposed.

Unless I hear a strong move to change the rule above, I actually

prefer

to stay with what we have.

/js

--
Juergen Schoenwaelder               International University Bremen
<http://www.eecs.iu-bremen.de/>       P.O. Box 750 561, 28725

Bremen,

Germany
--
!! This message is brought to you via the `ipv6mib' mailing list.
!! Please do not reply to this message to unsubscribe. To unsubscribe

or

adjust
!! your settings, send a mail message to

<[EMAIL PROTECTED]>

!! or look at https://www.ibr.cs.tu-bs.de/mailman/listinfo/ipv6mib.




--------------------------------------------------------------------
IETF IPv6 working group mailing list
[EMAIL PROTECTED]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to