Re: [mpls] Some following questions about ISD

["Dongjie (Jimmy)" <[e-mail]>]

Hi Tony,

From: Tony Li <[e-mail]> On Behalf Of Tony Li
Sent: Sunday, September 17, 2023 1:28 AM
To: Dongjie (Jimmy) <[e-mail]>
Cc: [e-mail]; MPLS Working Group <[e-mail]>; [e-mail]
Subject: Re: [mpls] Some following questions about ISD


Hi Jimmy,


[Jie] Agree the order of NASes may cause semantic difference, while the encoding order may change depend on the RLD limitation (thus not align with the ingress node’s intended order).  Maybe the framework or solution draft could provide a guidance on the default processing order of multiple NASes in one packet, regardless of their encoding order in packet.


That is up to the solution document.  With respect to the current ISD draft, section 5.5 might also say something like:

   If a label stack contains multiple NASes, then they MUST be processed as if they were processed in the order that they appear in the label stack,
        subject to the restrictions in Section 7.

[Jie] Yes, some text like this could be added to section 5.5. The text may also allow some special cases where the order of NASes in the label stack was due to the constraint of RLD, while the processing order may be different from the encoding order.


That’s not what I said or what I meant.  The Select NAS needs to be after the forwarding label, no question. It needs to be before the next forwarding label.  It could be before or after a HBH NAS.

[Jie] This is clearer. Hope the text in section 7 and 9 of the mpls-mna-hdr draft could further clarifies the rules of NAS placement and processing (current text about NAS placement and processing are somehow mixed in section 7).


You should propose any clarifications to the mailing list.

[Jie] OK, I will prepare some text for these two sections.

-Jie

Tony