Re: [mpls] Comments on draft-gandhi-mpls-stamp-pw

Hi Rakesh,

Thanks for addressing my comments.
Glad to become your new co-author. :-)
Cheers,
Xiao Min

Original

From: RakeshGandhi <rgandhi.ietf@gmail.com>
To: 肖敏10093570;
Cc: draft-gandhi-mpls-stamp-pw@ietf.org <draft-gandhi-mpls-stamp-pw@ietf.org>;mpls@ietf.org <mpls@ietf.org>;
Date: 2024年02月10日 03:57
Subject: Re: [mpls] Comments on draft-gandhi-mpls-stamp-pw

Thanks Xiao Min for the detailed review.

We have posted a new revision with the suggested changes:
https://www.ietf.org/archive/id/draft-gandhi-mpls-stamp-pw-05.html
DIFF: https://author-tools.ietf.org/iddiff?url2=draft-gandhi-mpls-stamp-pw-05

Please see inline with <RG1>...

On Fri, Nov 3, 2023 at 8:16 AM Rakesh Gandhi <rgandhi.ietf@gmail.com> wrote:

Hi Xiao Min,

Thanks for your review comments. Please see replies inline with <RG>.

On Fri, Nov 3, 2023 at 2:57 AM <xiao.min2@zte.com.cn> wrote:

Dear Authors,

I've read draft-gandhi-mpls-stamp-pw-04 and found it's valuable.

<RG1> Great, thanks!

In Section 4 and 5, two different packet formats are defined for STAMP Session-Sender Test Packet and STAMP Session-Reflector Test Packet respectively, a list of PW types on the application of each packet format may be helpful, e.g., for Ethernet PW, I think the STAMP Test Packet without IP/UDP Header should be used.

<RG> Yes, that's a good suggestion, we can include Ethernet PW in the next revision.

<RG1> Added this in Section 3.2.

The idea is to require only one packet format for a specific PW type.

<RG> Yes, we could add a table for PW type and STAMP format to be used in the next revision. For example, the following PW could use STAMP packet without IP/UDP header.
MPLS Ethernet PW Data Traffic [RFC4448]

L2-Specific Sublayer (L2SS) used in L2TPv3 PW Data Traffic [RFC3931]

Private Line Emulation [draft-ietf-pals-ple] Data Traffic

TDM over IP [RFC5087] Data Traffic (no IP Header case)

 <RG1> Added this in Section 3.2.

Besides, RFC 5085 defines three types of VCCV Control Channel, it's helpful if this draft can provide some analysis on how and why a specific type of VCCV Control Channel is choosed for STAMP over PW encapsulation, and I believe fate sharing between the PW data packet and the STAMP-over-PW test packet is an important argument behind the selection.

<RG> Yes, that's a good point. There are three types defined in RFC 5085. The scope of the draft is type 3 (TTL expiry). We can add some text for other types for completeness.

<RG1> Added this in Section 3, last paragraph.

Although the current version of this document focuses on STAMP over PW only, I don't see a reason why STAMP over LSP can't be included, I believe one document covering both STAMP over PW and STAMP over LSP is preferable.

<RG> Yes, agree. We can include MPLS LSP in the next revision.

<RG1> Added this in Section 3.1.

Thanks,
Rakesh

thanks,
Rakesh

Best Regards,
Xiao Min

_______________________________________________
 mpls mailing list
 mpls@ietf.org
 https://www.ietf.org/mailman/listinfo/mpls