Gunter - I strongly support Option #2 and strongly support Ketan's recommendation that an MSD sub-type be used to advertise ERLD. This is the unified framework that the MSD advertisement has been designed to support.
The following documents provide a unified definition of this mechanism: https://datatracker.ietf.org/doc/draft-ietf-isis-segment-routing-msd/ https://datatracker.ietf.org/doc/draft-ietf-ospf-segment-routing-msd/ https://datatracker.ietf.org/doc/draft-ietf-idr-bgp-ls-segment-routing-msd/ (The last one needs a refresh.) If we can update the related ERLD documents to align I think we will have an excellent solution. (Note: in the case of https://datatracker.ietf.org/doc/draft-ietf-idr-bgp-ls-segment-routing-rld/ perhaps that can be combined with https://datatracker.ietf.org/doc/draft-ietf-idr-bgp-ls-segment-routing-msd/ - but I leave that to the respective authors to work out.) Les > -----Original Message----- > From: Lsr <[email protected]> On Behalf Of Van De Velde, Gunter (Nokia > - BE/Antwerp) > Sent: Wednesday, June 13, 2018 2:10 AM > To: Ketan Talaulikar (ketant) <[email protected]>; [email protected]; > [email protected]; > [email protected] > Subject: Re: [Lsr] Signalling ERLD (ISIS, OSPF and BGP-LS) > > It is desirable that same understanding of TLVs ([ELC, RLD] or [ERLD]) are > signaled for ISIS, OSPF and BGP-LS. > > If the WG's can manage to agree upon a decision (option1/2/3 or 4), then > next, have a look into how to encode the TLV so that we have a clean > technological solution space. > > G/ > > -----Original Message----- > From: Ketan Talaulikar (ketant) [mailto:[email protected]] > Sent: Wednesday, June 13, 2018 10:45 > To: Van De Velde, Gunter (Nokia - BE/Antwerp) > <[email protected]>; [email protected]; [email protected]; > [email protected] > Subject: RE: Signalling ERLD (ISIS, OSPF and BGP-LS) > > Hi Gunter, > > In that case, I concur with you that option (2) is better than the others. My > only difference in opinion is that ERLD not have its own separate TLV but > instead get advertised as a new MSD sub-type - it is just a different > encoding. > > Thanks, > Ketan > > -----Original Message----- > From: Van De Velde, Gunter (Nokia - BE/Antwerp) > <[email protected]> > Sent: 13 June 2018 13:55 > To: Ketan Talaulikar (ketant) <[email protected]>; [email protected]; > [email protected]; > [email protected] > Subject: RE: Signalling ERLD (ISIS, OSPF and BGP-LS) > > Indeed, the debate that made BGP-LS to go down the ERLD path is of > pragmatic motivation. > > The major Readable Label Depth use-case is entropy. Hence, if the ERLD TLV > is available, then ELC can be implicitly assumed. No pragmatic reason to > signal > separately, as it just make things more complex then should be. > > >From a holistic perspective having something similar, yet different, in both > IGP and BGP-LS encoding seems to make little sense and only bring > confusion (router/controller implementers and network operators). > > The ways to address this in IGP and BGP-LS going forward: > 1) do nothing and leave all as it is (it has potential to create massive > confusion) > 2) only signal ERLD TLV in IGP and BGP > 3) signal ELC TLV and RLD TLV (unclear pragmatic value of explicit signaling > of > ELC TLV compared to option (2)) > 4) signal ELC TLV, RLD TLV and ERLD TLV (it has all, but is much much more > complex as option (2)) > > I believe that option (2) is the best option: > * it bring the needed readable label depth value to operators > * most simple solution for implementers (routers and controller) > * easy to understand with no confusion > * is compliant with draft-ietf-mpls-spring-entropy-label-10 > > G/ > > -----Original Message----- > From: Ketan Talaulikar (ketant) [mailto:[email protected]] > Sent: Wednesday, June 13, 2018 08:05 > To: Van De Velde, Gunter (Nokia - BE/Antwerp) > <[email protected]>; [email protected]; [email protected]; > [email protected] > Subject: RE: Signalling ERLD (ISIS, OSPF and BGP-LS) > > Hi Gunter, > > The difference in IGP signalling seems to be because the ELC is a capability > which is advertised differently than ERLD which is a limit. Are you saying > that > ELC does not have value by itself without the ERLD? > > IMHO it makes sense to retain ELC as capability of the router (as specified in > the IGP specs) and position ERLD as a MSD sub-type for indicating the limit. > This way we have the flexibility of signalling ERLD both per node and per > ingress link/LC level. > > Thanks, > Ketan > > -----Original Message----- > From: Idr <[email protected]> On Behalf Of Van De Velde, Gunter (Nokia > - BE/Antwerp) > Sent: 12 June 2018 19:28 > To: [email protected]; [email protected]; [email protected] > Subject: [Idr] Signalling ERLD (ISIS, OSPF and BGP-LS) > > In LSR WG the following drafts document the signaling of ELC and RLD: > * draft-ietf-ospf-mpls-elc > * draft-ietf-isis-mpls-elc > > When exporting this information using BGP-LS encoding to a controller, there > is need for BGP-LS extension by means of new TLVs. > > BGP-LS is signaling ERLD (entropy capable readable label depth) ISIS/OSPF is > signaling individually ELC and RLD > > I was working upon the IANA section, and discovered some inconsistency > that should be addressed: > * Why is IGP signaling individual ELC and RLD? ERLD is what was decided upon > (https://tools.ietf.org/html/draft-ietf-mpls-spring-entropy-label-10) > * What are the plans to request IANA code points for these drafts? > * (E)RLD seems to have meaning only from NODE perspective, (I assume that > LINK ERLD is not of any value at all, is that a correct assumption?) > > G/ > > > -----Original Message----- > From: Idr [mailto:[email protected]] On Behalf Of internet- > [email protected] > Sent: Tuesday, June 12, 2018 15:25 > To: [email protected] > Cc: [email protected] > Subject: [Idr] I-D Action: draft-ietf-idr-bgp-ls-segment-routing-rld-02.txt > > > A New Internet-Draft is available from the on-line Internet-Drafts > directories. > This draft is a work item of the Inter-Domain Routing WG of the IETF. > > Title : Signalling ERLD using BGP-LS > Authors : Gunter Van de Velde > Wim Henderickx > Matthew Bocci > Keyur Patel > Filename : draft-ietf-idr-bgp-ls-segment-routing-rld-02.txt > Pages : 6 > Date : 2018-06-12 > > Abstract: > This document defines the attribute encoding to use for BGP-LS to > expose ERLD "Entropy capable Readable Label Depth" from a node to a > centralised controller (PCE/SDN). > > > > The IETF datatracker status page for this draft is: > https://datatracker.ietf.org/doc/draft-ietf-idr-bgp-ls-segment-routing-rld/ > > There are also htmlized versions available at: > https://tools.ietf.org/html/draft-ietf-idr-bgp-ls-segment-routing-rld-02 > https://datatracker.ietf.org/doc/html/draft-ietf-idr-bgp-ls-segment-routing- > rld-02 > > A diff from the previous version is available at: > https://www.ietf.org/rfcdiff?url2=draft-ietf-idr-bgp-ls-segment-routing-rld- > 02 > > > 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. > > Internet-Drafts are also available by anonymous FTP at: > ftp://ftp.ietf.org/internet-drafts/ > > _______________________________________________ > Idr mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/idr > > _______________________________________________ > Idr mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/idr > > _______________________________________________ > Lsr mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/lsr _______________________________________________ spring mailing list [email protected] https://www.ietf.org/mailman/listinfo/spring
