Re: [Lsr] Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

[Weiqiang Cheng <[e-mail]>]

Thanks Aijun,

That is the point. 

Draft-lin provides a new dimension that supports algorithm flexibility under the same metric type. 

The both drafts are compatible.

 

B.R.

Weiqiang Cheng

 

 

发件人: Lsr [mailto:[e-mail]] 代表 Aijun Wang
发送时间: 2022年3月24日 09:29
收件人: 'Shraddha Hegde'; 'linchangwang'; 'Acee Lindem (acee)'; 'Chenmengxiao'; [e-mail]
抄送: [e-mail]
主题: Re: [Lsr] Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

 

Hi, Shraddha:

The metric-type based metric is different from the algorithm-based metric, I think we should not mixed them together.

Or, how you express the intention, that the link use different metric value for same metric type, but different Flex-Algo?

 

Best Regards

 

Aijun Wang

China Telecom

 

From: [e-mail] <[e-mail]> On Behalf Of Shraddha Hegde
Sent: Thursday, March 24, 2022 1:20 AM
To: linchangwang <[e-mail]>; Acee Lindem (acee) <acee=[e-mail]>; Chenmengxiao <[e-mail]>; Aijun Wang <[e-mail]>; [e-mail]
Cc: [e-mail]
Subject: Re: [Lsr] Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

 

Hi,

 

The draft https://www.ietf.org/archive/id/draft-ietf-lsr-flex-algo-bw-con-02.txt introduced generic metric

which has the metric-type and metric-value fields. It is possible that metric-type 128 and metric type 129 are advertised for the link and FAD definition for Flex-algo 128 uses metric-type 128 and flex-algo 129 uses metric-type 129.

It appears to me that the usecase you describe can be solved with generic metric.

 

Rgds

Shraddha

 

 

Juniper Business Use Only

From: Lsr <[e-mail]> On Behalf Of linchangwang
Sent: Wednesday, March 23, 2022 7:42 PM
To: Acee Lindem (acee) <acee=[e-mail]>; Chenmengxiao <[e-mail]>; Aijun Wang <[e-mail]>; [e-mail]
Cc: [e-mail]
Subject: Re: [Lsr] Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

 

[External Email. Be cautious of content]

 

 

Hi Acee

 

Very much appreciated- especially in the middle of the busy IETF week.

Thanks for your guidance with the text. We will  post  an update to the draft after the IETF week.

 

 

Best Regards

Changwang Lin

 

 

From: Lsr [mailto:[e-mail]] On Behalf Of Acee Lindem (acee)
Sent: Wednesday, March 23, 2022 5:31 PM
To: chenmengxiao (RD); Aijun Wang; [e-mail]
Cc: [e-mail]
Subject: Re: [Lsr] Re: Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

 

Hi Mengxiao, Aijun, 

 

Thanks for the clarification of the purpose of the and use case. It would be good to state that this document adds algorithm-specific link metrics as opposed to a dedicated metric in the title, abstract, and introduction. Will follow up but task multiplexing right now so it isn’t a good time to suggest text. 

 

Thanks,

Acee

 

From: Chenmengxiao <[e-mail]>
Date: Wednesday, March 23, 2022 at 12:26 AM
To: Aijun Wang <[e-mail]>, Acee Lindem <[e-mail]>, "[e-mail]" <[e-mail]>
Cc: "[e-mail]" <[e-mail]>
Subject: 答复: [Lsr] Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

 

Hi, Aijun:

 

Thank you for the clarification of FAPM sub-TLV. Really helpful :-)

 

 

Hi, Acee:

 

Thank you for the review and comment.

 

why you couldn’t just use the algorithm-specific metric in section 8 and 9 and draft-ietf-lsir-flex-algo

>>>> About the FAPM sub-TLV, I agree with Aijun that it’s used for Prefix Metric, not for Link Metric. According to draft-ietf-lsr-flex-algo, it is usually used in multi-area and multi-domain scenarios. IMHO, FAPM may be not a good solution for our case.

 

It seems to me that the use case is fairly obscure

>>>> Sorry for that we did not make it clear in the draft. We have revised the use case in the slide, and will update the draft later (maybe after the meeting).

 

Case 1:

 

   A------C

   |      |

   |      |

   |      |

   B------D

 

We have two network slices for the traffic from A to D. For slice 1, the network operator expects to use A-B-D as the primary path and A-C-D as the backup path. For slice 2, A-C-D is the primary path and A-B-D is the backup path.

Bandwidth resources are reserved along the primary paths for slices. On the backup path, no dedicated resources are reserved, and the bandwidth is shared with BE traffics.

The metric-type of the two Flex-Algorithms are the same since they both care about the bandwidth resources.

For Flex-Algo 128, we hope that metrics of link A-B and B-D are smaller than link A-C and C-D. But for Flex-Algo 129, we hope link A-C and C-D have smaller metrics.

 

Case 2:

 

    A------C

   /|*     |

  / |  *   |

 /  |    * |

E---B------D

 

There is TE-tunnel (or SR Policy) between A and D. A uses the tunnel as a short-cut path to D.

In Flex-Algo 128 and 129, forward adjacency is enabled for the tunnel A-D to allow other nodes to see it.

The metric of tunnel A-D is different in Flex-Algo 128 and 129 (for example, the physical path in Flex-Algo 128 is A-B-D, in Flex-Algo 129 is A-C-D).

So, we hope A can advertise the virtual link A-D with different metrics for Flex-Algo 128 and 129.

 

 

Best Regards

 

Mengxiao Chen

 

 

发件人: Lsr [mailto:[e-mail]] 代表 Aijun Wang
发送时间: 2022年3月23日 9:22
收件人: 'Acee Lindem (acee)' <acee=[e-mail]>; [e-mail]
抄送: [e-mail]
主题: Re: [Lsr] Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

 

Hi, Acee:

 

The Sub-TLV described in Section 8 and section 9 of the draft-ietf-lsr-flex-algo is for Prefix Metric, not for Link Metric.

Such sub-TLV is carried with the prefix advertisement. Will you change the name and application scope of such Sub-TLV?

 

Currently, the name is “IS-IS Flexible Algorithm Prefix Metric Sub-TLV”, its application scope is limited in:

“The IS-IS FAPM Sub-TLV is a sub-TLV of TLVs 135, 235, 236, and 237”.  

 

Same situation as OSPF

 

Best Regards

 

Aijun Wang

China Telecom

 

From: [e-mail] <[e-mail]> On Behalf Of Acee Lindem (acee)
Sent: Wednesday, March 23, 2022 2:02 AM
To: [e-mail]
Cc: [e-mail]
Subject: [Lsr] Question on "Advertisement of Dedicated Metric for Flexible Algorithm in IGP"

 

Speaking as WG member:

 

Hi Co-authors,

 

I’m read this draft and I really don’t see why you couldn’t just use the algorithm-specific metric in section 8 and 9 and draft-ietf-lsir-flex-algo? It seems to me that the use case is fairly obscure and there is nothing to prevent usage of these metrics for this use case and the draft is simply a matter of semantics. 

 

Thanks,

Acee

-------------------------------------------------------------------------------------------------------------------------------------
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!