[spring] operator requirements for draft-srcompdt-spring-compression-requirement

[Martin Horneffer <[e-mail]>]

Dear srcomp dt, and spring wg,

thanks a lot for the enormous effort to collect and describe all the 
requirements for compression mechanisms, and for already starting the 
analysis! A true work of merit.

 From an operator’s point of view I would like to add two requirements 
that I believe to be crucial to any kind of new overarching 
architecture: address management and address aggregation.

In my eyes, SRv6 does have the great potential to allow a new 
architectures that span many still separate network domains (access, 
aggregation, backbone, service areas, etc) and greatly simplify and 
streamline their operation. However, in order to allow this in an 
already existing operator environment, I really see these two points as 
essential.

This probably has already been covered by Dirk’s mail below, and I tried 
to make the point during the IETF109 session, but probably wasn’t clear 
enough. I haven’t seen any discussion of it yet.


In any case I tried to find a wording that might be suitable for 
addition to the requirements document. There are of course many ways to 
tackle the issue, and this is just one attempt. Please let me know what 
you think of it.

Best regards,
Martin


Proposed additions for draft-srcompdt-spring-compression-requirement:

4.3.4.  Number Resource Management

    Description: The compression mechanism SHOULD fit into existing IPv6 
address structures. It SHOULD NOT require management of a new kind of 
number resource that needs to be coordinated for all network domains 
that potentially could be connected to each other. Network domains 
rather take existing parts of their address space to provide their local 
functions, while staying fully interoperable with all other domains.

    Rationale: In larger organizations different network domains (e.g. 
access, aggregation, backbone, service areas) are managed by different 
organizational units. Number resources such as IP addresses and numeric 
identifiers must be organized in a way, that a) makes sure that every 
network domain gets enough resources (e.g. address space) to meet its 
needs, and b) conflicts between domains are prevented. Network operators 
have solved this problem for resources such as IPv4 and IPv6 addresses 
already and can relatively easily base new technology on this. On the 
other hand introducing new types of number resources might impose 
serious costs on all affected organizational units and thus seriously 
impede the introduction of the related technology.

    Metric: The compression mechanism fits into existing IPv6 address 
structures. It does not require management of a new kind of number 
resource that needs to be coordinated for all network domains that are 
potentially involved.

4.3.5.  Address Aggregation

    Description: The compression mechanism MUST support address 
aggregation between network domains. It SHOULD support address 
aggregation within a domain.

    Rationale: In larger organizations with multiple network domains and 
related organizational units it is effectively impossible to exactly 
foresee and plan the accumulated scale requirements for any reasonable 
future. Domain overarching architectures will fail if they do not apply 
serious aggregation of addresses at least at the borders between network 
domains.

    Metric: The compression mechanism allows address aggregation at 
least between network domains, and at as many additional levels as possible.





Am 19.11.20 um 16:46 schrieb Dirk Steinberg:
> Hello SPRING WG,
> 
> I have read the SRComp design team requirements draft
> and would like to comment.
> 
> I truly believe that a SID compression scheme MUST integrate into the
> existing SRv6 framework. Otherwise it does not make much sense,
> or said another way, it will not be a SID compression scheme for SRv6
> at all but another animal altogether.
> 
> SID compression should be used where the use case justifies it, i.e.
> strict path TE inside a given domain. Inter-Domain usage of SRv6,
> especially end systems in data centers, may have different requirements
> and thus decide to use uncompressed SRv6 SIDs. It is important that
> a SID list that describes a service that spans across multiple domains
> be able to contain both compressed and uncompressed SIDs.
> Consequently, the same CP needs to support both compressed and
> uncompressed SIDs.
> 
> I am currently working on an architecture based on SRv6 for different
> domains within a carrier network. These domains have different
> requirements and also different hardware capabilities that may lead
> to different designs for each subnetwork. But all these domains/
> subnetworks must be able to interoperate seamlessly based on SRv6
> standards, regardless of whether SID compression is used or not.
> 
> Therefore I strongly agree that Appendix A should be part of the draft.
> 
> I would also like to suggest another requirement:
> IMHO the single biggest advantage that SRv6 has compared to
> MPLS is aggregation (route summarization), something that is
> absolutely not possible with MPLS labels (SIDs).
> Aggregation (CIDR) is the very technology that has enabled the Internet
> to scale and to become the worldwide internetwork that it is today.
> In retrospect I believe the omission of aggregation has been the
> biggest design mistake in MPLS -- but back then there were a lot
> of other factors and the idea to use a very short tag for forwarding.
> After all Tag Switching and MPLS were inspired from ATM
> and within this context aggregation made no sense.
> 
> Consequently I propose to add to the draft the requirement that the
> SID compression scheme MUST be compatible with aggregation,
> i.e. it must be possible to express the reachability of a given set of
> SIDs (maybe in some domain or data center) using a summary prefix.
> 
> Thanks and Cheers
> Dirk
> 
> 
> 
> 
> 
> 
> On Thu, Nov 19, 2020 at 3:21 PM Ahmed Bashand <[e-mail] 
> <mailto:[e-mail]>> wrote:
> 
>     I also agree that the requirements in Appendix A should be part of
>     the draft. Having of existing standard as a basis greatly simplifies
>     the development and deployment of any compression scheme
> 
> 
>     Thanks
> 
> 
>     Ahmed
> 
> 
> 
>     On 11/19/20 12:58 AM, Ran Pang(联通集团中国联通研究院-本部) wrote:
>>     Hi Weiqiang and WG,
>>     I read the draft and agree with the requirements specified in it.I
>>     think the requirements in Appendix A should be part of the draft
>>     in the next version.
>>         China Unicom is working on a network evolution plan for SRv6
>>     now,  and we have done some field trials based on SRv6. In order
>>     to maintain the continuity of  the functionality, we suggest the
>>     solution based on the SRv6 standards.
>>
>>     Best regards,
>>     Pang Ran
>>
>>         *From:* 程伟强 <mailto:[e-mail]>
>>         *Date:* 2020-11-15 23:27
>>         *To:* spring <mailto:[e-mail]>
>>         *CC:* srcomp <mailto:[e-mail]>; spring-chairs@ietf.o
>>         <mailto:[e-mail]>
>>         *Subject:* [spring] Fw:New Version Notification for
>>         draft-srcompdt-spring-compression-requirement-01.txt
>>
>>         Hi Group,
>>
>>         SR compression design team have submitted a new version of
>>         compression requirement draft.
>>
>>         Main changes as follows:
>>
>>         - added 3 items about scalibility with agreement within the
>>         design team
>>
>>         - added an appendix including 3 items without without
>>         unanimous consensus within the design team
>>
>>         - some minor text issue fixed
>>
>>         Please review it and let us know your comments.
>>
>>
>>         BTW: We will have 1-hour session for the design team topic on
>>         Friday and welcome to join us.
>>
>>
>>         B.R.
>>
>>         Weiqiang on behalf of design team
>>
>>
>>         ----邮件原文----
>>         发件人：internet-drafts <[e-mail]>
>>         <mailto:[e-mail]>
>>         收件人：Weiqiang Cheng <[e-mail]>
>>         <mailto:[e-mail]>,Sander Steffann
>>         <[e-mail]> <mailto:[e-mail]>,SJM Steffann
>>         <[e-mail]> <mailto:[e-mail]>
>>         抄　送: (无)
>>         发送时间：2020-11-15 22:58:57
>>         主题：New Version Notification for draft-srcompdt-spring-
>>         compression-requirement-01.txt
>>
>>
>>         A new version of I-D, draft-srcompdt-spring-compression-requirement-01.txt
>>         has been successfully submitted by Weiqiang Cheng and posted to the
>>         IETF repository.
>>
>>         Name: draft-srcompdt-spring-compression-requirement
>>         Revision: 01
>>         Title: Compressed SRv6 SID List Requirements
>>         Document date: 2020-11-13
>>         Group: Individual Submission
>>         Pages: 13
>>         URL:
>>         https://www.ietf.org/archive/id/draft-srcompdt-spring-compression-requirement-01.txt
>>         <https://www.ietf.org/archive/id/draft-srcompdt-spring-compression-requirement-01.txt>
>>         Status:
>>         https://datatracker.ietf.org/doc/draft-srcompdt-spring-compression-requirement/
>>         <https://datatracker.ietf.org/doc/draft-srcompdt-spring-compression-requirement/>
>>         Htmlized:
>>         https://datatracker.ietf.org/doc/html/draft-srcompdt-spring-compression-requirement
>>         <https://datatracker.ietf.org/doc/html/draft-srcompdt-spring-compression-requirement>
>>         Htmlized:
>>         https://tools.ietf.org/html/draft-srcompdt-spring-compression-requirement-01
>>         <https://tools.ietf.org/html/draft-srcompdt-spring-compression-requirement-01>
>>         Diff:
>>         https://www.ietf.org/rfcdiff?url2=draft-srcompdt-spring-compression-requirement-01
>>         <https://www.ietf.org/rfcdiff?url2=draft-srcompdt-spring-compression-requirement-01>
>>
>>         Abstract:
>>          This document specifies requirements for solutions to compress SRv6
>>          SID lists.
>>
>>
>>
>>
>>         Please note that it may take a couple of minutes from the time of submission
>>         until the htmlized version and diff are available at
>>         tools.ietf.org <http://tools.ietf.org>.
>>
>>         The IETF Secretariat
>>
>>
>>
>>         Subject：New Version Notification for draft-srcompdt-spring-
>>         compression-requirement-01.txt
>>
>>
>>         A new version of I-D, draft-srcompdt-spring-compression-requirement-01.txt
>>         has been successfully submitted by Weiqiang Cheng and posted to the
>>         IETF repository.
>>
>>         Name: draft-srcompdt-spring-compression-requirement
>>         Revision: 01
>>         Title: Compressed SRv6 SID List Requirements
>>         Document date: 2020-11-13
>>         Group: Individual Submission
>>         Pages: 13
>>         URL:
>>         https://www.ietf.org/archive/id/draft-srcompdt-spring-compression-requirement-01.txt
>>         <https://www.ietf.org/archive/id/draft-srcompdt-spring-compression-requirement-01.txt>
>>         Status:
>>         https://datatracker.ietf.org/doc/draft-srcompdt-spring-compression-requirement/
>>         <https://datatracker.ietf.org/doc/draft-srcompdt-spring-compression-requirement/>
>>         Htmlized:
>>         https://datatracker.ietf.org/doc/html/draft-srcompdt-spring-compression-requirement
>>         <https://datatracker.ietf.org/doc/html/draft-srcompdt-spring-compression-requirement>
>>         Htmlized:
>>         https://tools.ietf.org/html/draft-srcompdt-spring-compression-requirement-01
>>         <https://tools.ietf.org/html/draft-srcompdt-spring-compression-requirement-01>
>>         Diff:
>>         https://www.ietf.org/rfcdiff?url2=draft-srcompdt-spring-compression-requirement-01
>>         <https://www.ietf.org/rfcdiff?url2=draft-srcompdt-spring-compression-requirement-01>
>>
>>         Abstract:
>>          This document specifies requirements for solutions to compress SRv6
>>          SID lists.
>>
>>
>>
>>
>>         Please note that it may take a couple of minutes from the time of submission
>>         until the htmlized version and diff are available at
>>         tools.ietf.org <http://tools.ietf.org>.
>>
>>         The IETF Secretariat
>>
>>
>>
>>     如果您错误接收了该邮件，请通过电子邮件立即通知我们。请回复邮件到
>>     [e-mail] <mailto:[e-mail]>，即可以退
>>     订此邮件。我们将立即将您的信息从我们的发送目录中删除。 If you have
>>     received this email in error please notify us immediately by
>>     e-mail. Please reply to [e-mail]
>>     <mailto:[e-mail]> ,you can unsubscribe from this
>>     mail. We will immediately remove your information from send
>>     catalogue of our.
>>
>>     _______________________________________________
>>     spring mailing list
>>     [e-mail]  <mailto:[e-mail]>
>>     https://www.ietf.org/mailman/listinfo/spring  <https://www.ietf.org/mailman/listinfo/spring>
>     _______________________________________________
>     spring mailing list
>     [e-mail] <mailto:[e-mail]>
>     https://www.ietf.org/mailman/listinfo/spring
>     <https://www.ietf.org/mailman/listinfo/spring>
> 
> 
> _______________________________________________
> spring mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/spring
>