
Message-Id: <v02140b1daef1ee3c6225@[171.69.128.114]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Thu, 2 Jan 1997 15:57:45 -0800
To: David Perkins <dperkins@scruznet.com>
Sender:iesg-request@ietf.org
From: Fred Baker <fred@cisco.com>
Subject: Re: Problems with Frame Relay DTE MIB
Cc: cbrown@cadia.com

At 4:31 PM 12/23/96, David Perkins wrote:
>Hi
>
>I just ran the MIB from draft-ietf-iplpn-frmib-dte-09.txt through the
>latest copy of SMICng. It flagged the following items:
>
>The following indices do not specify a range: frDlcmiIfIndex,
>frCircuitIfIndex, and frErrIfIndex. This is not the problem with
>the FR DTE MIB, but with the definition of TC InterfaceIndex from
>the IF MIB. Fred - help me get this problem fixed in the IF MIB.

I would suggest that you make your comment to the Interface MIB WG.

By the way, can you advise me why it did not complain when I did the same
thing? I am using the May 1996 version, and it didn't complain. Could it
possibly be that there is no requirement in SNMP for INTEGERs to have
specified ranges? I know that this is a bugaboo of yours (at one time, you
wanted to restrict interface indices to 1..65535), but I don't believe that
the specifications require that.

>The trap, frDLCIStatusChange, is not specified in a notification
>group. Thus, implementation requirements for the trap is ambiguous.
>Note also that group frTrapGroup is specified as optional in the
>module compliance specification. Thus, there is no requirement
>to implement the objects. This is bad for interoperablity. Either
>the objects must be removed from the MIB module, or the compliance
>specification changed to indicate in which situations it is
>mandatory to implement the objects.

It would have been helpful if SMICng had made this observation.

Also, This has been a couple of years in development (the lack of progress
being for several reasons, not the least of which is that there is no
burning reason to change RFC 1315, this is being done to go to Draft
Standard, not to meet an operational need), and went through last calls on
several lists - one of which was the IETF list. It seems that I have some
history with you coming in very late on MIBs and wanting changes long after
the last call has expired. Would it be too terribly awful to ask you to
make your comments sometime before the horse is out the gate?

>Regards,
>/dperkins@scruznet.com, David T. Perkins
>Date: 12/23/96, Time: 16:31:56