Re: [Pals] [mpls] draft-zzhang-intarea-generic-delivery-functions

["Jeffrey (Zhaohui) Zhang" <[e-mail]>]

Hi Adrian,

  *   But as John says, lots of hardware then (and probably now) sees bottom of stack and starts to sniff the next nibble.
Can I assume you are talking about multiple label stacks next to each other?

At least in the proposed GDF use case, multiple label stacks are separated by GDF headers, who always start with 0000 nibble.
That is similar to the following:

  *   A BIER header separates the <base tunnel label + BIER label (BoS)> stack and the label stack at the beginning of BIER payload
  *   A MPLS network built on PWs that are realized by another infrastructure MPLS network
Jeffrey
From: Kireeti Kompella <kireeti=[e-mail]>
Sent: Thursday, February 25, 2021 6:19 PM
To: [e-mail]; John E Drake <[e-mail]>; 'Stewart Bryant' <[e-mail]>
Cc: 'mpls' <[e-mail]>; [e-mail]; Ron Bonica <[e-mail]>; [e-mail]; [e-mail]; Jeffrey (Zhaohui) Zhang <[e-mail]>; Kireeti Kompella <[e-mail]>
Subject: Re: [mpls] draft-zzhang-intarea-generic-delivery-functions

[External Email. Be cautious of content]

Hi Adrian,

I'm all for multiple label stacks (if meaningful).  I had put that in an early version of the FAI bSPL, with a "real end stack" bit for those woke forwarding engines that needed to know.

The hard lesson I take away from those FEs that look at the nibble following the EoS is to make sure it isn't 0x4 or 0x6, not so much that a single label stack is the only viable solution.  So, keep this option alive ....

Cheers,
Kireeti.
From: Adrian Farrel <[e-mail]<mailto:[e-mail]>>
Date: Thursday, February 25, 2021 at 13:53
To: 'John E Drake' <jdrake=[e-mail]<mailto:jdrake=[e-mail]>>, 'Stewart Bryant' <[e-mail]<mailto:[e-mail]>>
Cc: 'mpls' <[e-mail]<mailto:[e-mail]>>, [e-mail]<mailto:[e-mail]> <[e-mail]<mailto:[e-mail]>>, Kireeti Kompella <[e-mail]<mailto:[e-mail]>>, Ron Bonica <[e-mail]<mailto:[e-mail]>>, [e-mail]<mailto:[e-mail]> <[e-mail]<mailto:[e-mail]>>, [e-mail]<mailto:[e-mail]> <[e-mail]<mailto:[e-mail]>>, 'Jeffrey (Zhaohui) Zhang' <zzhang=[e-mail]<mailto:zzhang=[e-mail]>>
Subject: RE: [mpls] draft-zzhang-intarea-generic-delivery-functions
> Multiple labels, each with the BoS set, was suggested during the MPLS-TP days
> (Niel Harrison, in particular, was a big proponent) but was shot down because it
> would apparently break existing hardware implementations.

Yeah, I was a big fan of that. I even thought it was architecturally the right thing to do for carrying one MPLS service over an MPLS transport. It would have solved many of the MPLS-TP requirements.

But as John says, lots of hardware then (and probably now) sees bottom of stack and starts to sniff the next nibble. Thus, contiguous label stacks and control words were the only viable solution. It took George Swallow a while to beat this into my head.

Cheers,
Adrian


Juniper Business Use Only
From: Stewart Bryant <[e-mail]<mailto:[e-mail]>>
Sent: Tuesday, February 23, 2021 10:36 AM
To: John E Drake <[e-mail]<mailto:[e-mail]>>
Cc: Stewart Bryant <[e-mail]<mailto:[e-mail]>>; Jeffrey (Zhaohui) Zhang <zzhang=[e-mail]<mailto:zzhang=[e-mail]>>; Rakesh Gandhi <[e-mail]<mailto:[e-mail]>>; mpls <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Kireeti Kompella <[e-mail]<mailto:[e-mail]>>; Ron Bonica <[e-mail]<mailto:[e-mail]>>; <[e-mail]<mailto:[e-mail]>> <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>
Subject: Re: draft-zzhang-intarea-generic-delivery-functions

[External Email. Be cautious of content]

Having an ACH channel type per set of structures, or a BoS label value per set of structures would be far more consistent with the MPLS model.

Stewart

On 23 Feb 2021, at 15:27, John E Drake <[e-mail]<mailto:[e-mail]>> wrote:

Hi,
Why wouldn't we just use GAL/G-ACh for this as it is already a published RFC that was designed for things like this?   I.e., we could use a GAL and define families of G-ACh codepoints which describe exactly what is in the area between the MPLS label stack and the payload.  E.g.,  G-Ach for IOAM:  [value #1 = hop by hop IOAM w/ no subsequent metadata, value #2 = end to end IOAM w/ no subsequent metadata, value #3 = hop by hop IOAM w/ subsequent metadata, value #4 = end to end IOAM w/ subsequent metadata].  The subsequent metadata would then have the same basic structure.
Yours Irrespectively,

John


Juniper Business Use Only
From: mpls <[e-mail]<mailto:[e-mail]>> On Behalf Of Jeffrey (Zhaohui) Zhang
Sent: Tuesday, February 23, 2021 10:19 AM
To: Stewart Bryant <[e-mail]<mailto:[e-mail]>>; Rakesh Gandhi <[e-mail]<mailto:[e-mail]>>
Cc: mpls <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Kireeti Kompella <[e-mail]<mailto:[e-mail]>>; Ron Bonica <[e-mail]<mailto:[e-mail]>>; <[e-mail]<mailto:[e-mail]>> <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>
Subject: Re: [mpls] draft-zzhang-intarea-generic-delivery-functions

[External Email. Be cautious of content]

Hi Stewart,

GDFH has both a "this header" (for a particular function) and a "next header". One GDFH can point to another GDFH, who could also point to another header (e.g. MPLS).

IOAM could be integrated into GDF. I understand that this is just a thought and we have not concluded on that yet, but with that thought the headers would be like the following:

   Transport label + GDFH label (BoS) + GDFH (for IOAM) + GDFH (for other functions not already provided by PW itself) + PW label + PW payload or G-Ach

Depending on the situation, the GDFH (for IOAM) and GDFH (for other functions) could be swapped. The PW label is essentially another label stack - the earlier stack stopped at the GDFH label.

Now what if IOAM is not integrated into GDF? I see the following in its -06 revision:
    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    | IOAM Indicator Label                  | TC  |1|  TTL          |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+<-+
    |0 0 0 1|Version| Reserved      | IOAM G-ACh                    |  |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+  |
    | Reserved      | Block Number  | IOAM-OPT-Type |IOAM HDR Length|  |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+  I
    |                                                               |  O
    |                                                               |  A
    ~                 IOAM Option and Data Space                    ~  M
    |                                                               |  |
    |                                                               |  |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+<-+
    |0 0 0 1|Version| Reserved      | Channel Type                  |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |                                                               |
    |                                                               |
    ~                 Payload + Padding                             ~
    |                                                               |
    |                                                               |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

     Figure 5: Example MPLS Encapsulation with Another G-ACh with IOAM

I am not sure if that is a good idea:


  *   RFC 5586 says "The G-ACh MUST NOT be used to transport user traffic". To me IOAM traffic is user traffic with OAM information.
  *   It puts one G-Ach header after another G-Ach header. As far as I understand it, G-Ach does not have a "next header" concept and using 0001b to indicate that another G-Ach header follows is not safe.
  *   In case PW payload, the above will put the PW label before the IOAM label. Having those extra IOAM/G-Ach headers between the PW+IOAM label and PW payload adds complexity to the fast path processing.

Jeffrey

From: Stewart Bryant <[e-mail]<mailto:[e-mail]>>
Sent: Monday, February 22, 2021 7:30 PM
To: Jeffrey (Zhaohui) Zhang <[e-mail]<mailto:[e-mail]>>
Cc: Stewart Bryant <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; mpls <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Kireeti Kompella <[e-mail]<mailto:[e-mail]>>; Ron Bonica <[e-mail]<mailto:[e-mail]>>; <[e-mail]<mailto:[e-mail]>> <[e-mail]<mailto:[e-mail]>>
Subject: Re: draft-zzhang-intarea-generic-delivery-functions

[External Email. Be cautious of content]

What happens if an operator wants to run both iOAM and GDFH at the same time and the packet is a PW packet?

What does the packet look like and how does the forwarder know what to do?

- Stewart

On 22 Feb 2021, at 22:49, Jeffrey (Zhaohui) Zhang <[e-mail]<mailto:[e-mail]>> wrote:

Hi Stewart,

This thread started with your comment "Please see the note that I sent about iOAM who also want to sit after BoS ... and both of you want the same space that PALS and DetNet is already using", but now it seems that we're on the same page - GDFH starting with 0000b is fine and is not competing with IOAM or PW/DETNET CW?

Thanks.
Jeffrey

From: Stewart Bryant <[e-mail]<mailto:[e-mail]>>
Sent: Monday, February 22, 2021 5:15 AM
To: Jeffrey (Zhaohui) Zhang <[e-mail]<mailto:[e-mail]>>
Cc: Stewart Bryant <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; mpls <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Kireeti Kompella <[e-mail]<mailto:[e-mail]>>; Ron Bonica <[e-mail]<mailto:[e-mail]>>; <[e-mail]<mailto:[e-mail]>> <[e-mail]<mailto:[e-mail]>>
Subject: Re: draft-zzhang-intarea-generic-delivery-functions

[External Email. Be cautious of content]

The DetNet CW is described in RFC8964 and is





      0                   1                   2                   3

      0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1

     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

     |0 0 0 0|                Sequence Number                        |

     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+



                       Figure 5: DetNet Control Word



In all defined control words



The 0000 is simply ECMP defeat and has no other purpose.



0001 means ACH



An ACH is currently defined not to carry service/user data - it is a control/OAM channel.



You cannot assume anything about a payload starting 0000.



In MPLS the bottom label (alone) defines how you process the payload. So you know that you have a CW from the bottom label and by no other means.



In other words the the FEC of the bottom label and its associated parameters are the way that signalling protocol knows what instructions to give the forwarder, and the way that the forwarder knows what to do with the packet is from the instructions associated with the BoS label. This is the universal model for MPLS including for IP packets.



Stewart



On 19 Feb 2021, at 15:42, Jeffrey (Zhaohui) Zhang <[e-mail]<mailto:[e-mail]>> wrote:

Hi Stewart,

I still have to read more about DetNet, but I am not sure if there is a real contention with PALS.

My understanding of 0000 nibble in PW control world is that it is only to prevent a transit node from mistaking the payload as IP. Is it supposed to indicate that any payload starting with 0000 is PW payload? I hope not.

Use of 0000 nibble in GDFH is also just to prevent transit nodes from mistaking it as IP. It does indicate it is GDFH. It should be able to co-exist with PW CW.

Thanks.
Jeffrey

-----Original Message-----
From: Jeffrey (Zhaohui) Zhang
Sent: Thursday, February 18, 2021 10:35 PM
To: Stewart Bryant <[e-mail]<mailto:[e-mail]>>
Cc: [e-mail]<mailto:[e-mail]>; mpls <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Kireeti Kompella <[e-mail]<mailto:[e-mail]>>; Ron Bonica <[e-mail]<mailto:[e-mail]>>; <[e-mail]<mailto:[e-mail]>> <[e-mail]<mailto:[e-mail]>>
Subject: RE: draft-zzhang-intarea-generic-delivery-functions

Stewart, all,

I apologize for not responding to this in time. I some how accidentally moved a few wg mailing list email folders to a place where I could not see so I missed all the discussions.
Let me catch up all the emails and then reply.

Thanks.
Jeffrey

-----Original Message-----
From: Stewart Bryant <[e-mail]<mailto:[e-mail]>>
Sent: Tuesday, January 12, 2021 9:59 AM
To: Jeffrey (Zhaohui) Zhang <[e-mail]<mailto:[e-mail]>>
Cc: Stewart Bryant <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; mpls <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Kireeti Kompella <[e-mail]<mailto:[e-mail]>>; Ron Bonica <[e-mail]<mailto:[e-mail]>>; <[e-mail]<mailto:[e-mail]>> <[e-mail]<mailto:[e-mail]>>
Subject: Re: draft-zzhang-intarea-generic-delivery-functions

[External Email. Be cautious of content]


Thank you Jeffery

Please see the note that I sent about iOAM who also want to sit after BoS ... and both of you want the same space that PALS and DetNet is already using.

We plan to have a joint session on this hosted by PALS at the next IETF, but I think we also need to include the iOAM people.

This has scope to get very messy as we find new candidates for BoS metadata so we really need to take a holistic position to ensure the future health the MPLS protocol.

- Stewart

On 12 Jan 2021, at 14:27, Jeffrey (Zhaohui) Zhang <[e-mail]<mailto:[e-mail]>> wrote:

Hi,

I just posted https://urldefense.com/v3/__https://datatracker.ietf.org/doc/draft-zzhang-intarea-generic-delivery-functions/__;!!NEt6yMaO-gk!QyBnufJO58LP6Diq96EdYEe2kxFtiItOdNuXbu_RIMekK2pkpOj4Mmj7b9MseV-Y$<https://urldefense.com/v3/__https:/datatracker.ietf.org/doc/draft-zzhang-intarea-generic-delivery-functions/__;!!NEt6yMaO-gk!QyBnufJO58LP6Diq96EdYEe2kxFtiItOdNuXbu_RIMekK2pkpOj4Mmj7b9MseV-Y$> .

The initial version was posted to the tsvwg (https://urldefense.com/v3/__https://tools.ietf.org/html/draft-zzhang-tsvwg-generic-transport-functions-00__;!!NEt6yMaO-gk!QyBnufJO58LP6Diq96EdYEe2kxFtiItOdNuXbu_RIMekK2pkpOj4Mmj7b5lS_Jea$<https://urldefense.com/v3/__https:/tools.ietf.org/html/draft-zzhang-tsvwg-generic-transport-functions-00__;!!NEt6yMaO-gk!QyBnufJO58LP6Diq96EdYEe2kxFtiItOdNuXbu_RIMekK2pkpOj4Mmj7b5lS_Jea$> ). After discussions/feedback we are re-homing it to intarea wg. This new version also contains quite some changes based on the comments and feedback that we received (special thanks to Stewart).

Comments and suggestions are appreciated.

Thanks.
Jeffrey

Juniper Business Use Only


Juniper Business Use Only


Juniper Business Use Only


Juniper Business Use Only

Juniper Business Use Only



Juniper Business Use Only


Juniper Business Use Only


Juniper Business Use Only