[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Comments on LMP



Baktha,
   Please see comments inline.
 
Thanks,
Jonathan
 
-----Original Message-----
From: Baktha Muralidharan [mailto:muralidb@cisco.com]
Sent: Friday, March 22, 2002 11:23 AM
To: Jonathan Lang
Cc: ccamp@ops.ietf.org
Subject: Comments on LMP

Hi Jonathan,

        Looks the following comments that I sent some time ago
        got lost.

Thanks,

/Baktha


X-Sender: muralidb@funnel.cisco.com
X-Mailer: QUALCOMM Windows Eudora Version 4.3.2
Date: Thu, 14 Mar 2002 23:02:27 -0500
To: Jonathan Lang <jplang@calient.net>
From: Baktha Muralidharan <muralidb@cisco.com>
Subject: Comments on LMP
Cc: ccamp@ops.ietf.org, muralidb@cisco.com

Hi Jonathan

         We have the following comments on LMP draft 3:
--------------------------------------------------
1. In the current LMP draft (draft-ietf-ccamp-lmp-03.txt), section 7 talks   about Message ID usage. It states,
"Unacknowledged messages sent with the MESSAGE_ID object   SHOULD be retransmitted until the message is acknowledged or until a retry limit is reached."

What are the subsequent actions that need to take place when the above conditions occur ? 
 
[Jonathan]If a retry limit is reached, the message must not be retransmitted    
 
 2. When Nacking a link summary, the current message format doesn't allow indicate what's wrong with each interface mapping.. "per-interface" error codes would be helpful. i.e. The TE link and each data link object in the Nack message   needs an error code to localize the failure. 
 
[Jonathan] From Section 13.7.3: "The DATA_LINK objects included in the LinkSummaryNack message MUST include accpetable values for all negotiable parameters.  If the LinkSummaryNack includes DATA_LINK Objects for non-negotiable parameters, they MUST be copied from the DATA_LINK Objects received in the LinkSummary message."
 
Why isn't this enough to indicate what's wrong with each interface mapping?
 
 3. When a channel status msg is received with message ID less than   an earlier (TE-Link) message, it might still be accepted per the draft. So, in the case where it is accepted, should the accepted message ID become the basis for all future message ID (validations)? 
 
[Jonathan]Message Id values are stored for each TE-Link as well as data channel.  The stored Message_Id values are only updated when a message arrives with a Message_Id value that is greater than the currently stored value.
 
4. CC ID can be numbered (i.e. IPv4 format) per OIF/OUNI. However, the current LMP draft does not support it. 
 
 [Jonathan] LMP defines the CCID as a 32-bit field.  For the in-band case of OIF/OUNI, an IPv4 address could be used.
 
 ------------------------------------------------------------
Thanks,

/Baktha Muralidharan
Cisco Systems, Inc.
Chelmsford, MA.