Sasha, Thanks for your time to respond my emails.
Regards, Vahid [email protected] > On Apr 14, 2019, at 5:19 PM, Alexander Vainshtein > <[email protected]> wrote: > > Vahid, > The SPRING WG mailing list is not a proper forum for discussing proprietary > tutorials. > They should rather be discussed with whoever has published them, > > Regarding your reference to Anycast-SID appearing as the first SID in the > list – well, it is a matter of interpretation. From my POV all SID lists > implicitly begin with some identification of the head-end node, so what you > see as the first SID in the list looks as the 2nd SID in the implicit list. > > > My 2c, > Sasha > > Office: +972-39266302 > Cell: +972-549266302 > Email: [email protected] > <mailto:[email protected]> > > From: vahid tavajjohi <[email protected] > <mailto:[email protected]>> > Sent: Sunday, April 14, 2019 2:16 PM > To: Alexander Vainshtein <[email protected] > <mailto:[email protected]>> > Cc: [email protected] <mailto:[email protected]> > Subject: Re: [spring] Anycast-SID > > Sasha, > Sorry, but I don’t get my answer. > Also, in SRTE document on “ > http://www.segment-routing.net/tutorials/2017-03-06-segment-routing-traffic-engineering-srte/ > > <http://www.segment-routing.net/tutorials/2017-03-06-segment-routing-traffic-engineering-srte/> > ", you can see that it uses Anycast-SID as first SID of SID list (Page 48). > > <image001.png> > > > Also, in Cisco affiliated website: “ > https://xrdocs.io/design/blogs/2018-05-09-metro-design-implementation-guide/ > <https://xrdocs.io/design/blogs/2018-05-09-metro-design-implementation-guide/> > “ you can see that they use Anycast prefix (on two nodes) with one SID, > without clearing N-flag clear. (violating RFC 8402 section 3.2) > > <image002.png> > > > And these things confused me for meaning of Anycast-SID and effect of > violating RFC8402 rules in network. > I think SPRING WG, should add more and specific explanation for Anycast > models and usages. > > > Regards, > Vahid > [email protected] <mailto:[email protected]> > > > On Apr 14, 2019, at 3:22 PM, Alexander Vainshtein > <[email protected] <mailto:[email protected]>> > wrote: > > Vahid, > Anycast-SID can be used in the middle of the list of SIDs defining a SR-TE > LSP. > > > Regards, > Sasha > > Office: +972-39266302 > Cell: +972-549266302 > Email: [email protected] > <mailto:[email protected]> > > From: vahid tavajjohi <[email protected] > <mailto:[email protected]>> > Sent: Sunday, April 14, 2019 1:02 PM > To: Alexander Vainshtein <[email protected] > <mailto:[email protected]>> > Cc: [email protected] <mailto:[email protected]> > Subject: Re: [spring] Anycast-SID > > Sasha, > RFC 8402 section 3.3.1 illustrated and mentioned that Group A members are > using anycast address 192.0.2.10/32 and the Anycast-SID 100. So, is N-flag > cleared? and how they used /32? > Also, “ draft-ietf-isis-segment-routing-extensions-23” section 2.1.1.2 states > that “ The router MUST ignore the N-Flag on a received Prefix-SID if the > prefix has a Prefix length different than /32 (IPv4) or /128 (IPv6)”. So, if > I use prefix greater than /32, routers ignores n-flag and there is no need to > clear n-flag by my self. > > I glad to clarify me for these. > > > > > On Apr 14, 2019, at 1:49 PM, Alexander Vainshtein > <[email protected] <mailto:[email protected]>> > wrote: > > Vahid, > RFC 8402 states that “IGP-Anycast segment MUST NOT reference a particular > node” while the /32 IPv4 address that iBGP advertises as the NH of VPN-IP > routes of course references a particular node, namely one the has allocated > the labels in these routes. > > I.e., if you clear N-Flag, you MUST NOT use the /32 IPv4 address in the > prefix as the BGP NH in any labeled routes advertised by this node. > > Regards, > Sasha > > Office: +972-39266302 > Cell: +972-549266302 > Email: [email protected] > <mailto:[email protected]> > > From: vahid tavajjohi <[email protected] > <mailto:[email protected]>> > Sent: Sunday, April 14, 2019 11:49 AM > To: Alexander Vainshtein <[email protected] > <mailto:[email protected]>> > Cc: [email protected] <mailto:[email protected]> > Subject: Re: [spring] Anycast-SID > > Sasha, > Ok it is clear. Based on your scenario, If I set Anycast-SID and clear N-Flag > , problem will solve? > > Regards, > Vahid > > > > > On Apr 14, 2019, at 12:25 PM, Alexander Vainshtein > <[email protected] <mailto:[email protected]>> > wrote: > > Vahid, > This is not about PCE, it is (eventually) about traffic. > May I suggest that you perform the following test: > 1. Define the same IPv4 /32 prefix in two nodes A and B as a Node-SID. > 2. Set up a BGP/MPLS IP VPN service that is represented in nodes A and > C, but not in Node B and that uses shortest path SR LSPs as tunnels. Take > care of iBGP in A using the IP address in question as the NH of VPN-IP > routes it advertises while not defining BGP in B. > 3. Run test traffic over this service with ingress in C and (expected) > egress in A. > > You will see that, this traffic will pass if Dist (C A) < Dist (C, B), and > will be blackholed if Dist (C, A) > Dist (C, B). And this is all you need to > know. > > > Regards, > Sasha > > Office: +972-39266302 > Cell: +972-549266302 > Email: [email protected] > <mailto:[email protected]> > > From: vahid tavajjohi <[email protected] > <mailto:[email protected]>> > Sent: Sunday, April 14, 2019 10:45 AM > To: Alexander Vainshtein <[email protected] > <mailto:[email protected]>> > Cc: [email protected] <mailto:[email protected]> > Subject: Re: [spring] Anycast-SID > > Sasha, > It is obvious that it violates rules, my point is what happens in the > network. I can’t find any clue in my LAB environment and any other documents. > I don’t clear N-FLAG in anycast but PCE uses Anycast-SID in SRTE. I clear > N-FLAG in anycast, PCE does not uses Anycast-SID but I tried SRTE with > explicit-path, it worked correctly. So, my point is there are no explanation > that describes effect of violating rules. > Also, definition of Anycast-SID is not clear. For example, is Anycast for > plane separation is different than Anycast for HA(ABR) or not? > > I hope my explanation is clear. > > Regards, > Vahid > > > > > > On Apr 14, 2019, at 11:55 AM, Alexander Vainshtein > <[email protected] <mailto:[email protected]>> > wrote: > > Vahid, > Section 2.1.1.2 of the IS-IS Extensions for Segment Routing > <https://tools.ietf.org/html/draft-ietf-isis-segment-routing-extensions-23> > draft defines N-Flag in the Prefix-SID Sub-TLV as following (the relevant > text is highlighted): > > N-Flag: Node-SID flag. If set, then the Prefix-SID refers to > the router identified by the prefix. Typically, the N-Flag is > set on Prefix-SIDs attached to a router loopback address. The > N-Flag is set when the Prefix-SID is a Node-SID as described in > [RFC8402 <https://tools.ietf.org/html/rfc8402>]. > > An RF C 8402 states in Section 3.2: > > An IGP Node-SID MUST NOT be associated with a prefix that is owned by > more than one router within the same routing domain. > > > Therefore, the answer to your first question “What happens if I don’t > implement these rules in network?” is simple: Your implementation violates a > mandatory requirement of the Segment Routing architecture. > > > Regards, > Sasha > > Office: +972-39266302 > Cell: +972-549266302 > Email: [email protected] > <mailto:[email protected]> > > -----Original Message----- > From: spring <[email protected] <mailto:[email protected]>> On > Behalf Of vahid tavajjohi > Sent: Saturday, April 13, 2019 7:39 AM > To: [email protected] <mailto:[email protected]> > Subject: [spring] Anycast-SID > > Hi SPRING WG, > > I have a question about Anycast-SID. > 1- In RFC 8402 section 3.2, it mentioned that “Node-SID MUST NOT be > associated with a prefix that is owned by more than one router within the > same routing domain”. Also, in section 3.3, it mentioned that “ An > IGP-Anycast segment MUST NOT reference a particular node”. > > 2- Also, we have N-Flag “isis-segment-routing-extensions-23” that indicates > whether Prefix-SID related to a node or not. > > My questions are: > 1- What happens if I don’t implement these rules in network? > 2- What happens if I set anycast-sid on multiple nodes, but I don’t set > "n-flag clear” under loopback configuration? > 3- Why "IGP-Anycast segment MUST NOT reference a particular node” ? > > > Best Regards, > Vahid > _______________________________________________ > spring mailing list > [email protected] <mailto:[email protected]> > https://www.ietf.org/mailman/listinfo/spring > <https://www.ietf.org/mailman/listinfo/spring> > > ___________________________________________________________________________ > > This e-mail message is intended for the recipient only and contains > information which is > CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have > received this > transmission in error, please inform us by e-mail, phone or fax, and then > delete the original > and all copies thereof. > ___________________________________________________________________________ > > > ___________________________________________________________________________ > > This e-mail message is intended for the recipient only and contains > information which is > CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have > received this > transmission in error, please inform us by e-mail, phone or fax, and then > delete the original > and all copies thereof. > ___________________________________________________________________________ > > > ___________________________________________________________________________ > > This e-mail message is intended for the recipient only and contains > information which is > CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have > received this > transmission in error, please inform us by e-mail, phone or fax, and then > delete the original > and all copies thereof. > ___________________________________________________________________________ > > > ___________________________________________________________________________ > > This e-mail message is intended for the recipient only and contains > information which is > CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have > received this > transmission in error, please inform us by e-mail, phone or fax, and then > delete the original > and all copies thereof. > ___________________________________________________________________________ > > > ___________________________________________________________________________ > > This e-mail message is intended for the recipient only and contains > information which is > CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have > received this > transmission in error, please inform us by e-mail, phone or fax, and then > delete the original > and all copies thereof. > ___________________________________________________________________________
_______________________________________________ spring mailing list [email protected] https://www.ietf.org/mailman/listinfo/spring
