Re: [bess] Section 8.5 of draft-ietf-bess-rfc7432bis

[Gyan Mishra <[e-mail]>]

Hi Satya

Backwards compatibility is critical and completely understand.

However let’s say all routers are upgraded to support 8584 and Type 1 HRW
and negotiate the AC-DF capability exchange codepoint P2P then would all
routers supporting type 1 automatically start using Type 1 HRW Alg?

Kind Regards

Gyan

On Mon, Mar 6, 2023 at 12:21 AM Satya Mohanty (satyamoh) <[e-mail]>
wrote:

> Thanks Jorge.
>
>
>
> Gyan, indeed all the three algorithms have their use-cases.
>
>
>
> As pointed out below in case of inconsistency (the “Alg type”) in the DF
> election extended community for a given ES, it was decided that the DF
> election must default to the modulo-based simply because modulo-based was
> proposed first and all existing deployments already had it, before the
> other two (Pref-based and HRW) came into existence.
> https://www.rfc-editor.org/rfc/rfc8584.html#section-2.2.1
>
>
>
> Thanks,
>
> --Satya
>
>
>
> *From: *BESS <[e-mail]> on behalf of Jorge Rabadan (Nokia) <
> [e-mail]>
> *Date: *Friday, March 3, 2023 at 11:06 PM
> *To: *Gyan Mishra <[e-mail]>
> *Cc: *mdodge <[e-mail]>, [e-mail] <[e-mail]>
> *Subject: *Re: [bess] Section 8.5 of draft-ietf-bess-rfc7432bis
>
> Hi Gyan,
>
>
>
> I agree with your two first statements, but I’m afraid I don’t agree with
> this one:
>
> “I would think HRW would be the new default algorithm used for DF election
> and would not be a knob as it fixes the major deficiencies with the modulus
> algorithm.”
>
>
>
> The modulo-based DF algorithm remains the default algorithm in case of
> inconsistency in the Ethernet Segment peers. It uses DF Alg 0, while HRW
> uses type 1 and Preference type 2. HRW is not obsoleting the modulo-based
> or anything like that. The three algorithms are widely used in networks,
> and while the default one has limitations, it is simple and works out in
> many networks.
>
>
>
> Thank you!
>
> Jorge
>
>
>
>
>
> *From: *Gyan Mishra <[e-mail]>
> *Date: *Friday, March 3, 2023 at 11:50 PM
> *To: *Jorge Rabadan (Nokia) <[e-mail]>
> *Cc: *Menachem Dodge <[e-mail]>, [e-mail] <[e-mail]>
> *Subject: *Re: [bess] Section 8.5 of draft-ietf-bess-rfc7432bis
>
>
>
> *CAUTION:* This is an external email. Please be very careful when
> clicking links or opening attachments. See http://nok.it/ext for
> additional information.
>
>
>
>
>
> Hi Jorge
>
>
>
> As defined in RFC 8584 AC-DF and HRW are independent of each other.
>
>
>
> RFC 8584 updates RFC 7432, however RFC 7432 bis does not update any aspect
> of RFC 8584 including HRW as you stated.
>
>
>
> I would think HRW would be the new default algorithm used for DF election
> and would not be a knob as it fixes the major deficiencies with the modulus
> algorithm.
>
>
>
> Thanks
>
>
>
> Gyan
>
> On Fri, Feb 10, 2023 at 2:01 PM Jorge Rabadan (Nokia) <
> [e-mail]> wrote:
>
> Hi Menachem,
>
>
>
> The way I see it, draft-ietf-bess-rfc7432bis will obsolete RFC7432, but it
> does not update or change RFC8584, so I don’t think
> draft-ietf-bess-rfc7432bis needs to repeat the aspects of RFC8584.
>
>
>
> In particular, the AC-DF, as the other capabilities defined in other
> documents, is a capability that can be turned on or off. While it is very
> useful in many cases, it actually needs to be disabled in some others,
> e.g., draft-ietf-bess-evpn-mh-pa-07
> <https://datatracker.ietf.org/doc/html/draft-ietf-bess-evpn-mh-pa-07>
>
>
>
> So the answer to your second question could be that the AC-DF capability
> should be configurable in the implementation, and used on all cases where
> issues with individual Attachment Circuits may create blackholes. However,
> it has to be disabled in case of port-active multi-homing.
>
>
>
> My 2 cents..
>
>
>
> Thanks.
>
> Jorge
>
>
>
>
>
> *From: *BESS <[e-mail]> on behalf of Menachem Dodge <
> [e-mail]>
> *Date: *Friday, February 10, 2023 at 8:10 AM
> *To: *[e-mail] <[e-mail]>
> *Subject: *[bess] Section 8.5 of draft-ietf-bess-rfc7432bis
>
> Some people who received this message don't often get email from
> [e-mail]. Learn why this is important
> <https://aka.ms/LearnAboutSenderIdentification>
>
> Hello All,
>
>
>
> In section 8.5 of draft-ietf-bess-rfc7432bis there is a reference to the
> Finite State Machine in section 2.1 of RFC 8584.
>
>
>
> However, in section 4 of RFC 8584 the AC Influenced DF Election Capability
> is described and it states that this updates Step 3 of the procedure for
> service carving of RFC 7432.
>
>
>
> Shouldn’t this AC-DF update be mentioned in the procedures of section 8.5
> of draft-ietf-bess-rfc7432bis?
>
>
>
> Is AC Influenced DF Election the preferred DF election method ?
>
>
>
>
>
> According to RFC 8584 “ The procedure discussed in this section is
> applicable to the DF
>
>    election in EVPN services [RFC7432 <https://datatracker.ietf.org/doc/html/rfc7432>] and EVPN VPWS [RFC8214 <https://datatracker.ietf.org/doc/html/rfc8214>]. “
>
>
>
>
>
>
>
> Thank you kindly.
>
>
>
> Best Regards,
>
>
>
> *Menachem Dodge**​**​**​**​*
>
> System Architect
>
> [image: signature_3305758272]
>
> [phone]
>
> [e-mail]
>
> follow us on Linkedin <https://www.linkedin.com/company/drivenets>
>
> www.drivenets.com
>
> [image: DriveNets Network Cloud]
> <https://get.drivenets.com/mwc-2023-schedule-a-meeting>
>
>
>
> _______________________________________________
> BESS mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/bess
>
> --
>
> <http://www.verizon.com/>
>
> *Gyan Mishra*
>
> *Network Solutions Architect *
>
> *Email [e-mail] <[e-mail]>*
>
> *M 301 502-1347*
>
>
>
-- 

<http://www.verizon.com/>

*Gyan Mishra*

*Network Solutions A**rchitect *

*Email [e-mail] <[e-mail]>*



*M 301 502-1347*