Skip to main content

Early Review of draft-ietf-rtgwg-srv6-egress-protection-09
review-ietf-rtgwg-srv6-egress-protection-09-opsdir-early-hares-2023-08-08-00

Request Review of draft-ietf-rtgwg-srv6-egress-protection-09
Requested revision 09 (document currently at 16)
Type Early Review
Team Ops Directorate (opsdir)
Deadline 2023-08-05
Requested 2023-07-11
Requested by Yingzhen Qu
Authors Zhibo Hu, Huaimo Chen , Mehmet Toy , Chang Cao , Tao He
I-D last updated 2023-08-08
Completed reviews Rtgdir Early review of -11 by Tal Mizrahi (diff)
Intdir Early review of -10 by Bob Halley (diff)
Opsdir Early review of -09 by Susan Hares (diff)
Comments
Kindly request early reviews of this document, specifically focusing on its consistency and effectiveness in relation to existing mechanisms.
Assignment Reviewer Susan Hares
State Completed
Request Early review on draft-ietf-rtgwg-srv6-egress-protection by Ops Directorate Assigned
Posted at https://mailarchive.ietf.org/arch/msg/ops-dir/84gRkfYFaqfGj5KpqGwOb7oiKA8
Reviewed revision 09 (document currently at 16)
Result Has issues
Completed 2023-08-08
review-ietf-rtgwg-srv6-egress-protection-09-opsdir-early-hares-2023-08-08-00
This is an OPS-DIR review of draft-ietf-rtgwg-srv6-egress-protection-09.
This review should be taken as an early review just as other reviews.

Status: Serious Issues + editorial issues
Why: The text describes mechanisms, ISIS sub-TLVs, OSPF sub-TLVs, and security
issues. Each of these sections has problems that need to be fixed before WG LC.

What should the authors do:
I'll happily work with the authors to correct the text. The text in version 09
does not provide a clear way for interoperable implementations. I will also
review -12 to check if these issues have been resolved.

Mechanisms:
The mechanisms provide steps that are unclear and are not easily aligned with
the example. The mechanism section contains the following: - description of
figure 1 - step 1: a normal path set-up (PEA) - step 2: a backup pat set-up
(PEB)
    step 2a: PEB announcement to PEA with PEB
    step 2b: PEA processes the announcement and sends forwarding behavior to PEB
    step 2c: PEB processes forwarding based on Mirror SID into table
    step 2d: processing of anycast that links PEA/PEB to compute LFA with SIDs
- step 3: Steps in fail-over in failure scenario
   step 3a: failure detected via BFD
   step 3b: send IPv6 packet with an encapsulated packet (H.Encaps) at PL1
   step 3c: decapsulate IPv6 packet with End.M (a variant of End.DT6)
- step 4: Handle processing and forwarding a PEB

If these are the correct steps, then the example in section 3.2.
If these are not the correct steps, you confirm the reading is difficult.

Extensions to IS-IS.
The following fields need to be given a range.
Section 4.1
For figure 3 description, please correct the following:
a) length: Please give a valid range.
b) Flags: Normally a reserved value has a default setting (all zeros?) for
transmit,
       and ignore upon reception.
c) SRv6 Endpoint Function: Please give valid range.
d) SID: 16 octets (please give valid ranges.  For example, is SID 0 valid?)

for Figure 4, please valid ranges for length field.

Section 4.2
For figure 5, please correct the following:
length: Please give the valid range.
flags: Normally a reserved value has a default setting (all zeros?) for
transmit,
       and ignore upon reception.
SRv6 Endpoint Function: Please give valid range.
SID: 16 octets (please give valid ranges.  For example, is SID 0 v

The Security Considerations section should provide reasons why this mechanism
to provide re-routing does not provide a potential threat.  It may work in a
walled garden, but it is not clear there are "no extra security issues."