Re: [trill] New draft: draft-mizrahi-trill-loss-delay-00

[Tal Mizrahi <[e-mail]>]

Hi Gagan,

Thanks for your comments.
Please see my responses inline.

Tal.

From: Gagan Mohan Goel [mailto:[e-mail]]
Sent: Tuesday, February 19, 2013 10:24 AM
To: Tal Mizrahi
Cc: [e-mail]; Donald Eastlake ([e-mail]); Samer Salam (ssalam) ([e-mail]); [e-mail]
Subject: RE: New draft: draft-mizrahi-trill-loss-delay-00

Hi Tal,

Following are my views on the draft http://tools.ietf.org/html/draft-mizrahi-trill-loss-delay.

1) Section 4.2.3 SLR Message Reception

<snip>    The receiver computes the one-way packet delay with respect to a measurement interval. A measurement interval includes a sequence of
   1SLM message. The one-way packet delay is performed by comparing the counter values TXp and RXp at the beginning of the measurement
   interval, and the counter values TXc and RXc at the end of the  measurement interval (Figure 2):
<snip>

I understand that receiver can only calculate the one way packet loss (and not packet delay) with incoming SLM message. I think the above can be modified to:



 The receiver computes the one-way packet loss with respect to a measurement interval. A measurement interval includes a sequence of
   SLM message. The one-way packet loss is performed by comparing the counter values TXp and RXp at the beginning of the measurement
   interval, and the counter values TXc and RXc at the end of the  measurement interval (Figure 2):

[TM] Right. It is indeed a typo.

2) 4.1.2<http://tools.ietf.org/html/draft-mizrahi-trill-loss-delay-00#section-4.1.2>. 1SLM Message Reception

<snip> The receiver MUST maintain a reception counter for each peer MEP and test ID. Upon receiving a 1SLM packet, the receiver MUST verify that:



   o The 1SLM packet is destined to the current MEP.

   o The packet's MD level matches the MEP's MD level.

<snip>
May be another criteria that can be added to handle any transient network loop scenario:
Ø  Discard the message if counter received in the incoming 1SLM message is lower than current RX count for same testID.
[TM] Since packets can in some cases be received out-of-order, this filter may discard 1SLM packets that should be counted.

Regards
Gagan Mohan Goel

From: [e-mail] [mailto:[e-mail]] On Behalf Of Tal Mizrahi
Sent: Monday, February 18, 2013 8:17 PM
To: [e-mail]
Cc: [e-mail]; Donald Eastlake ([e-mail]); Samer Salam (ssalam) ([e-mail])
Subject: [trill] New draft: draft-mizrahi-trill-loss-delay-00

Hi,

We have uploaded a new draft titled "Loss and Delay Measurement in TRILL".

http://tools.ietf.org/html/draft-mizrahi-trill-loss-delay

Comments will be welcome.

Thanks,
Tal.