Re: [mpls] New Version Notification for draft-li-mpls-mna-entropy-00.txt

Tony Li <tony.li@tony.li> Tue, 07 February 2023 06:46 UTC

Sender: Tony Li <tony1athome@gmail.com>
Content-Type: text/plain; charset="utf-8"
Mime-Version: 1.0 (Mac OS X Mail 16.0 \(3731.300.101.1.3\))
From: Tony Li <tony.li@tony.li>
In-Reply-To: <202302071125088240525@zte.com.cn>
Date: Mon, 06 Feb 2023 22:45:48 -0800
Cc: draft-li-mpls-mna-entropy@ietf.org, mpls@ietf.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <B2E89728-AA7F-4E0E-A3AF-C0B7EA4EB635@tony.li>
References: <202302061608518696322@zte.com.cn,> <3406CB51-A701-498E-AA41-7271E91D23C4@tony.li> <202302071125088240525@zte.com.cn>
To: liu.yao71@zte.com.cn
Archived-At: <https://mailarchive.ietf.org/arch/msg/mpls/2m7xN600z5Nsjcc7U1vrxL9NXok>
Subject: Re: [mpls] New Version Notification for draft-li-mpls-mna-entropy-00.txt
Precedence: list

Hi Yao,

That’s exactly what I’m proposing.

T


> On Feb 6, 2023, at 7:25 PM, <liu.yao71@zte.com.cn> <liu.yao71@zte.com.cn> wrote:
> 
> Hi Tony,
> 
> If the choice is left to the encoding node, for the same network action, i.e, entropy, there may be two different encoding types defined for it? 
> 
> Yao
> 
> 
> ------------------Original------------------
> From: TonyLi <tony.li@tony.li>
> To: 刘尧00165286;
> Cc: draft-li-mpls-mna-entropy@ietf.org <draft-li-mpls-mna-entropy@ietf.org>;mpls@ietf.org <mpls@ietf.org>;
> Date: 2023年02月07日 01:50
> Subject: Re: [mpls] New Version Notification for draft-li-mpls-mna-entropy-00.txt
> 
> Hi,
> I consider encoding efficiency (and thus on-the-wire efficiency) to be paramount.
> If an entropy value is to be encoded with a number of other flag based actions, then that might have less overhead.
> Alternately, if a smaller entropy value is to be encoded, then it might be more efficient to have that in the same LSE as an opcode.
> The choice is left to the node that is encoding the sub-stack.
> Tony
>> On Feb 6, 2023, at 12:08 AM, liu.yao71@zte.com.cn wrote:
>> 
>> Hi Authors,
>> 
>> In this draft a new opcode is defined for entropy action, considering that both definig new opcode and leveraging opcode 3(Flag-based NA with AD) can carry in-stack NA with data, why the former method is chosen in this draft?
>> 
>> Thanks,
>> Yao

[mpls] Fwd: New Version Notification for draft-li… Tony Li
Re: [mpls] Fwd: New Version Notification for draf… liu.yao71
Re: [mpls] New Version Notification for draft-li-… Tony Li
Re: [mpls] New Version Notification for draft-li-… liu.yao71
Re: [mpls] New Version Notification for draft-li-… Tony Li