Hi Robert,

Thank you for your comments.
Please find my answers inline marked with Mengxiao>.

Thanks,
Mengxiao Chen

From: Robert Raszuk <[email protected]>
Sent: Thursday, July 14, 2022 8:28 PM
To: chenmengxiao (RD) <[email protected]>
Cc: [email protected]; linchangwang (RD) <[email protected]>; lihao (02566, 
RD) <[email protected]>
Subject: Re: [Lsr] FW: New Version Notification for 
draft-lin-lsr-srv6-service-sid-00.txt

Hi,

Thank you for publishing this document.

I have few comments about it.

#1 - While you are defining IGP extensions I think this document still belongs 
in BESS WG not in LSR WG.
Mengxiao> The solutions proposed in BESS are all BGP enabled. In this document 
we are discussing the solution without BGP. So, I don't quite understand why it 
belongs in BESS.

#2 - Injecting IPv4 sites to your IPv6 only IGP is IMO a terrible idea. BGP 
overlays are used to keep core IGPs stable and lean. Networks which do not want 
to run BGP in the core just use dual stack.
Mengxiao> Dual stack is certainly a solution without BGP. However, there are 
some scenarios that the core network is newly built and the operator only wants 
to deploy IPv6. To interconnect IPv4 islands under those scenarios, we usually 
need to build ipv4 over ipv6 tunnels (for example, GRE-tunnel) between the edge 
routers of different islands. This work is usually done manually, and the 
number of tunnels may be large. Then, the edge routers establish IGP neighbors 
through those tunnels. With the advertisement of SRv6 End.DT4 SIDs in IGP, the 
interconnection of IPv4 islands can be automatic and efficient. Besides, this 
mechanism may be extended to IPv6 only networks. For example, in the campus 
network not deploying BGP, advertising SRv6 End.DT6 SIDs along with color in 
IGP may be used to provide SRv6 TE service for IPv6 sites.

#3 However I do see value of general idea as expressed in this draft but only 
for deployment scenario where you DO NOT inject prefixes from sites to your IGP 
core and you use CSC (carrier's carrier) approach - still no BGP in IPv6 core. 
That means IPv4 sites exchange IPv4 reachability over the top and encapsulate 
packets to the other edge of IPv4 islands.
Mengxiao> That is an interesting idea. As mentioned above, the traditional 
solution builds IPv4 over IPv6 tunnels through which the edge routers of 
different islands establish IGP neighbors. So, the tunnel-based solution is a 
kind of CSC approach. Do you mean we can use SRv6 Service SIDs somehow to 
improve the tunnel-based solution for IGP? Or, we just use SRv6 Service SIDs to 
build tunnels instead of GRE?

So for #3 I would in fact state that this is not a bad idea. But not for #2 as 
currently written.

Many thx,
Robert.


On Thu, Jul 14, 2022 at 1:58 PM Chenmengxiao 
<[email protected]<mailto:[email protected]>> wrote:
Hi WG,

We posted a new draft: draft-lin-lsr-srv6-service-sid-00. It's about 
advertising SRv6 service SID in IGP.

For the IPv6 backbone networks not deploying BGP, for example, the campus 
network using IS-IS or OSPFv3, there may be requirements to interconnect IPv4 
islands. SRv6 Service SIDs like End.DT4 may be used to realize such 
requirements. Similar with the BGP based L3 Service over SRv6, SRv6 Service 
SIDs like End.DT4 may be used to realize such requirements. This document 
extends IS-IS and OSPFv3 to advertise SRv6 Service SIDs along with prefixes. 
The receiving router creates forwarding entries associated with SID.
In the IPv4 over IPv6 scenario, the IPv4 packets will be encapsulated in an 
outer IPv6 header with the destination address of SRv6 Service SID and 
forwarded according to the SID. When they are forwarded to the destination, the 
SID's owner decapsulates the outer IPv6 header and performs IPv4 table lookup 
to forward the inner IPv4 packet. Furthermore, it may be extended to support 
SRv6-TE Services in IGP.

Any questions or comments are welcomed.

Thanks,
Mengxiao Chen

-----Original Message-----
From: [email protected]<mailto:[email protected]> 
<[email protected]<mailto:[email protected]>>
Sent: Monday, July 11, 2022 4:38 PM
To: linchangwang (RD) 
<[email protected]<mailto:[email protected]>>; lihao (02566, 
RD) <[email protected]<mailto:[email protected]>>; chenmengxiao (RD) 
<[email protected]<mailto:[email protected]>>
Subject: New Version Notification for draft-lin-lsr-srv6-service-sid-00.txt


A new version of I-D, draft-lin-lsr-srv6-service-sid-00.txt
has been successfully submitted by Mengxiao Chen and posted to the IETF 
repository.

Name:draft-lin-lsr-srv6-service-sid
Revision:00
Title:IS-IS and OSPFv3 Extensions to Advertise SRv6 Service SID
Document date:2022-07-11
Group:Individual Submission
Pages:9
URL:            
https://www.ietf.org/archive/id/draft-lin-lsr-srv6-service-sid-00.txt
Status:         https://datatracker.ietf.org/doc/draft-lin-lsr-srv6-service-sid/
Htmlized:       
https://datatracker.ietf.org/doc/html/draft-lin-lsr-srv6-service-sid


Abstract:
   The IPv6 backbone networks only deploying IGP may be required to
   interconnect IPv4 islands. SRv6 Service SIDs like End.DT4 may be
   used to realize such requirements. This document extends IS-IS and
   OSPFv3 to advertise SRv6 Service SIDs.




The IETF Secretariat


-------------------------------------------------------------------------------------------------------------------------------------
本邮件及其附件含有新华三集团的保密信息,仅限于发送给上面地址中列出
的个人或群组。禁止任何其他人以任何形式使用(包括但不限于全部或部分地泄露、复制、
或散发)本邮件中的信息。如果您错收了本邮件,请您立即电话或邮件通知发件人并删除本
邮件!
This e-mail and its attachments contain confidential information from New H3C, 
which is
intended only for the person or entity whose address is listed above. Any use 
of the
information contained herein in any way (including, but not limited to, total 
or partial
disclosure, reproduction, or dissemination) by persons other than the intended
recipient(s) is prohibited. If you receive this e-mail in error, please notify 
the sender
by phone or email immediately and delete it!
_______________________________________________
Lsr mailing list
[email protected]<mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/lsr
_______________________________________________
Lsr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lsr

Reply via email to