Hi Lizhong, Apologies for the delay. I missed the email as it got landed up in different folder. Please see below,
[Lizhong] It would be better if NVO3 tunnel could be verified. And I believe NVO3 would require the tool to achieve that. <Nagendra>Am I assuming right that you intend to perform some validation at the terminating point (endpoint) with VXLAN (or NVGRE) header in such ping packet?. I am working on submitting a new version. I will consider this after discussing with other authors. Please let me know if my above assumption on your query is different. [Lizhong] I mean you could not check the dataplane validation of L2/L3 VN entry. I am not sure if it is useful for such validation. <Nagendra> I believe this would be equivalent to current validation machinery in MPLS (or other overlay based) network. Per my understanding, L2/L3 VN end point may be in different domains (multi tenant scenarios) where the NVE and the actual VN endpoints will be different domains and we don't need to leak any ping packets to the VN end points. But I think it is a good point to mention for clarity. Once again Thanks for your comments and happy to discuss more on the same. Thanks, Nagendra On 12/5/13 10:12 PM, "Lizhong Jin" <[email protected]> wrote: >> >1. the NVO3 ping uses 3503 as the destination port, then the transport >> >path of NVO3 Ping would be different with NVO3 tunnel (I am referring >> >to VXLAN tunnel), which means NVO3 ping could not detect the NVO3 >> >tunnel dataplane failure. Do you want to only detect the IP layer >> >failure for the overlay network? >> <Nagendra> You are right that the NVO3 ping uses a different destination >> port. But it also provides a way to perform ECMP pat validation. So when >> there are multiple paths between the ingress/egress, it provides a way >>to >> validate all paths. >[Lizhong] It would be better if NVO3 tunnel could be verified. And I >believe >NVO3 would require the tool to achieve that. > >> >> >2. The purpose of this draft is to validate the L2/L3 VN context. But >> >as my understanding, you could only detect the existence of the L2/L3 >> >VN context. >> >The validation of L2/L3 VN context would not only include the existence >> >of the entry, but also include the correctness of the output. >> >Maybe I have missed something important in the draft. >> <Nagendra> This is an initial draft where we just mentioned the L2/L3 VN >ID. >> It not only provides a way to check the presence of the VN but can be >> extended to carry relevant information like MAC that can be used by the >> egress to perform the validation. We will include the details and TLV in >future >> version. >[Lizhong] I mean you could not check the dataplane validation of L2/L3 VN >entry. I am not sure if it is useful for such validation. > >> >> Please let me know if this answers your query. > >Thanks >Lizhong > >> >> Thanks, >> Nagendra >> >> > >> >Regards >> >Lizhong >> > >> > >> >> -----Original Message----- >> >> From: Nagendra Kumar Nainar (naikumar) [mailto:[email protected]] >> >> Sent: 2013年12月4日 7:41 >> >> To: [email protected] >> >> Cc: Carlos Pignataro (cpignata) >> >> Subject: [nvo3] FW: I-D Action: draft-kumar-nvo3-overlay-ping-00.txt >> >> >> >> Hi, >> >> >> >> Below is the draft we published on NVO3 path ping. Please read the >> >> same and share your comments. >> >> >> >> Thanks, >> >> Nagendra >> >> >> >> >> >> > >> >> >A New Internet-Draft is available from the on-line Internet-Drafts >> >> >directories. >> >> > >> >> > >> >> > Title : Detecting NVO3 Overlay Data Plane failures >> >> > Author(s) : Nagendra Kumar >> >> > Carlos Pignataro >> >> > Dhananjaya Rao >> >> > Filename : draft-kumar-nvo3-overlay-ping-00.txt >> >> > Pages : 14 >> >> > Date : 2013-07-15 >> >> > >> >> >Abstract: >> >> > This document describes a simple and efficient mechanism to >>perform >> >> > L2 or L3 VN Context validation and to detect any data plane >failures >> >> > in IPv4 or IPv6 based overlay network providing L2 or L3 >virtualized >> >> > network. >> >> > >> >> > >> >> >The IETF datatracker status page for this draft is: >> >> >https://datatracker.ietf.org/doc/draft-kumar-nvo3-overlay-ping >> >> > >> >> >There's also a htmlized version available at: >> >> >http://tools.ietf.org/html/draft-kumar-nvo3-overlay-ping-00 >> >> > >> >> > >> >> >Internet-Drafts are also available by anonymous FTP at: >> >> >ftp://ftp.ietf.org/internet-drafts/ >> >> > >> >> >_______________________________________________ >> >> >I-D-Announce mailing list >> >> >[email protected] >> >> >https://www.ietf.org/mailman/listinfo/i-d-announce >> >> >Internet-Draft directories: http://www.ietf.org/shadow.html or >> >> >ftp://ftp.ietf.org/ietf/1shadow-sites.txt >> >> >> > >> > > > _______________________________________________ nvo3 mailing list [email protected] https://www.ietf.org/mailman/listinfo/nvo3
