Re: [pim] pim wg adoption call for draft-chen-pim-srv6-p2mp-path-01

["Ehsan Hemmati (ehemmati)" <[e-mail]>]

I do agree with Jeffrey and Object to adoption for the same reason.

Ehsan

On Jan 22, 2021, at 6:19 AM, Jeffrey (Zhaohui) Zhang <zzhang=[e-mail]<mailto:zzhang=[e-mail]>> wrote:

With IR, you don't need any of the following:

1. encoding (sub-)tree in the packet
2. core routers supporting the new forwarding behavior (i.e. new ASIC)

I am less concerned with edge devices having to support new forwarding behavior because they typically are capable of doing that w/o new ASIC; but core routers are more difficult with that, so we need to be careful with solutions requiring new forwarding behavior.

You mentioned 10x replication with draft-chen. That's 10 leaves but you also need to encode replication points. As I mentioned in the other email, it's good to see a concrete illustration of the SRv6 header.

Back to my original comment - how does it compare to BIER-TE, and do we really need another solution that requires new  ASIC on core routers.

Jeffrey

-----Original Message-----
From: pim <[e-mail]<mailto:[e-mail]>> On Behalf Of Xiejingrong (Jingrong)
Sent: Thursday, January 21, 2021 8:54 PM
To: Stig Venaas <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>
Subject: Re: [pim] pim wg adoption call for draft-chen-pim-srv6-p2mp-path-01

[External Email. Be cautious of content]


Hi all,

I have read the document and the presentation slides.
I think it is a clear proposal to solve some clear problems of existing approaches.

Comparing to IR, this proposal can do a stateless replication (say 10x replication) in 1 packet, thus increase the scalability of IR to 10X.
It can be seen as an extension to IR (a 1x replication), and the cost to scale to 10x replication is very deterministic and smooth.
It can be combined with IR seamlessly.
For example, IR could support a stateless 100x replication from an Ingress router to 100 Egress routers.
Then, this proposal could support a stateless 1000x replication using the combination of IR * SR-P2MP.

Contrary to the concern of scalability, I think this proposal has a linear scalability improvement comparing to IR-only, and therefore I support the adoption of this document.

To authors, it may be useful to clarify the scope and add some text about scalability consideration in the document.

Regards,
Jingrong


-----Original Message-----
From: pim [mailto:[e-mail]] On Behalf Of Stig Venaas
Sent: Tuesday, January 12, 2021 3:54 AM
To: [e-mail]<mailto:[e-mail]>
Subject: [pim] pim wg adoption call for draft-chen-pim-srv6-p2mp-path-01

Hi all

We discussed the draft draft-chen-pim-srv6-p2mp-path-01 in the last wg meeting. We did do an adoption call in the meeting, but we also would like to get input on the mailing list.

Please respond by January 25th whether you believe this document should be adopted.
The document can be found at
https://urldefense.com/v3/__https://tools.ietf.org/html/draft-chen-pim-srv6-p2mp-path-01__;!!NEt6yMaO-gk!X-bfbdJ7YKP527XztGdWD_d7FiiEJ138T4nEbEwJB0s-KGZNWzndp6TBzStPkbf-$

Regards,
Stig

_______________________________________________
pim mailing list
[e-mail]<mailto:[e-mail]>
https://urldefense.com/v3/__https://www.ietf.org/mailman/listinfo/pim__;!!NEt6yMaO-gk!X-bfbdJ7YKP527XztGdWD_d7FiiEJ138T4nEbEwJB0s-KGZNWzndp6TBzQoYlD8B$

_______________________________________________
pim mailing list
[e-mail]<mailto:[e-mail]>
https://urldefense.com/v3/__https://www.ietf.org/mailman/listinfo/pim__;!!NEt6yMaO-gk!X-bfbdJ7YKP527XztGdWD_d7FiiEJ138T4nEbEwJB0s-KGZNWzndp6TBzQoYlD8B$

Juniper Business Use Only

_______________________________________________
pim mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/pim