Re: PBB E-VPN and TRILL

[Sam Aldrin <[e-mail]>]

Donald,

From what I remember, at the WG session, the question of Trill interconnect is part of the charter for trill WG session or not, arose. The answer was, no, it was not part of the charter. (I do not see minutes to reference to, so basing it on my memory). We did not split hairs whether it is trill over x or trill campus extension, at that time. If you say, interconnect of Trill(ex: draft-aldrin-trill-data-center-interconnect)  is  ok to to be part of trill WG or dependent underlying technology group, then I am fine. Just wanted to get clarified.

Sam

Sent from my iPad

On May 10, 2012, at 12:38 PM, Donald Eastlake <[e-mail]> wrote:

> On Thu, May 10, 2012 at 3:25 PM, Sam Aldrin <[e-mail]> wrote:
>> Hi John,
>> 
>> Thanks for the reply.
>> 
>> The way I understood the below sentence from the charter, when I first read, was that it includes extending TRILL by interconnecting TRILL campuses or sites. But during last WG session, it wasn't so and was contentious topic. As TRILL is L2 technology and interconnecting them falls into L2VPN charter/territory.
> 
> I do not agree that TRILL is L2 technology.
> 
>> Which WG owns the charter for TRILL interconnect is not a major point, IMO (others may differ)
>> As long it is clear to the TRILL WG on where to publish the drafts related to interconnect and get reviewed/adopted, it is good. As of now, it is not clear.
> 
> So what was wrong with my statement that "TRILL over X" drafts should
> be in the TRILL WG or the X WG while it is OK for drafts that provide
> limited control plane interaction between TRILL islands, like the
> current L2VPN draft, to be in L2VPN? ("TRILL over X" drafts are drafts
> that pretty much just tell you how to encode any TRILL Data or TRILL
> Control packet over a technology X link, along with whatever security
> or other precautions you should take, so as to obtain full peering and
> cause the RBridges on that link, and any other fully connected
> RBridges, to be part of one unified TRILL campus.)
> 
> Thanks,
> Donald
> =============================
>  Donald E. Eastlake 3rd   [phone] (cell)
>  155 Beaver Street, Milford, MA 01757 USA
>  [e-mail]
> 
>> cheers
>> -sam
>> On May 10, 2012, at 12:06 PM, John E Drake wrote:
>> 
>>> Sam,
>>> 
>>> I think all that Donald is saying is that TRILL can use a multiplicity of trunking technologies (from the current charter: "but also additional ways to extend and optimize TRILL for the properties of the networks on which it is deployed.").
>>> 
>>> Thanks,
>>> 
>>> John
>>> 
>>> Sent from my iPhone
>>> 
>>>> -----Original Message-----
>>>> From: [e-mail] [mailto:[e-mail]] On Behalf
>>>> Of Sam Aldrin
>>>> Sent: Thursday, May 10, 2012 11:18 AM
>>>> To: Donald Eastlake
>>>> Cc: [e-mail]; [e-mail]
>>>> Subject: Re: PBB E-VPN and TRILL
>>>> 
>>>> Donald,
>>>> 
>>>> At the WG session it was asked whether interconnecting TRILL was in
>>>> TRILL charter of not. The answer was no. So, will the charter be edited
>>>> to reflect what you are eluding to below?
>>>> 
>>>> Sam
>>>> 
>>>> Sent from my iPhone
>>>> 
>>>> On May 10, 2012, at 10:56 AM, Donald Eastlake <[e-mail]> wrote:
>>>> 
>>>>> On Thu, May 10, 2012 at 10:35 AM, Lucy yong <[e-mail]>
>>>> wrote:
>>>>>> There is another draft to be considered.
>>>>>> 
>>>>>> http://tools.ietf.org/html/draft-yong-trill-trill-o-mpls-01.txt
>>>>> 
>>>>> There are a number of drafts but I think it is useful to make a
>>>>> distinction between drafts that provide full connectivity between
>>>>> TRILL switches (RBridges), so that they merge into a single TRILL
>>>>> campus, and drafts that provide for data plane connectivity but
>>>>> limited control plane connectivity for fault isolation, etc.
>>>>> 
>>>>> TRILL switches are routers. Like IP routers they logically strip the
>>>>> link envelope from TRILL frames they receive and add a new, possible
>>>>> different technology, link envelope on TRILL frames they send. Just as
>>>>> it is possible to have an IP routed region where none of the
>>>>> connections between IP routers is Ethernet but could be, for example,
>>>>> PPP, it is possible to have a TRILL campus where none of the
>>>>> connections between the RBridges is Ethernet but they are all some
>>>>> other technology X, such as PPP. So, drafts that are "TRILL over X"
>>>>> that are talking about full TRILL data and control plan connectivity
>>>>> don't really seem to me to have much to do with L2VPN. Such drafts
>>>>> should be done in either the TRILL WG or in the WG that does
>>>>> technology X.
>>>>> 
>>>>> It seems to me that L2VPN has to do with method that provide limited
>>>>> data plane interconnection for better fault isolation.
>>>>> 
>>>>> Thanks,
>>>>> Donald
>>>>> =============================
>>>>> Donald E. Eastlake 3rd   [phone] (cell)
>>>>> 155 Beaver Street, Milford, MA 01757 USA  [e-mail]
>>>>> 
>>>>>> Lucy
>>>>>> 
>>>>>> -----Original Message-----
>>>>>> From: [e-mail] [mailto:[e-mail]] On
>>>>>> Behalf Of Balaji Venkat Venkataswami
>>>>>> Sent: Thursday, May 10, 2012 2:34 AM
>>>>>> To: [e-mail]; [e-mail]
>>>>>> Cc: [e-mail]; [e-mail]
>>>>>> Subject: PBB E-VPN and TRILL
>>>>>> 
>>>>>> Hi  Giles,
>>>>>> 
>>>>>> I totally agree on this point. We have a solution that was presented
>>>>>> to the TRILL working group as well on the interconnection of TRILL
>>>>>> islands. We would like to concur with this and request for entering
>>>> our draft as well into the solution space.
>>>>>> 
>>>>>> The URL for the TRILL draft is as follows.
>>>>>> http://tools.ietf.org/html/draft-balaji-trill-over-ip-multi-level-05
>>>>>> 
>>>>>> 
>>>>>> thanks and regards,
>>>>>> balaji venkat
>>>>>> 
>>>>>> Sent: Wednesday, May 09, 2012 6:04 PM
>>>>>> To: [e-mail]
>>>>>> Cc: Ali Sajassi (sajassi); Stewart Bryant
>>>>>> Subject: PBB E-VPN and TRILL
>>>>>> 
>>>>>> Ali presented the PBB-EVPN draft at IETF83 in Paris:
>>>>>> 
>>>>>> http://tools.ietf.org/html/draft-ietf-l2vpn-pbb-evpn-01
>>>>>> 
>>>>>> This has since been updated:
>>>>>> 
>>>>>> http://tools.ietf.org/html/draft-ietf-l2vpn-pbb-evpn-02
>>>>>> 
>>>>>> During our discussions Ali mentioned that the authors would like the
>>>> TRILL section of the draft to be separated out into a separate draft.
>>>> In response Stewart questioned whether interconnecting TRILL islands is
>>>> in-scope for L2VPN.
>>>>>> 
>>>>>> Stewart, Nabil and I have just been discussing these issues.  Whilst
>>>> TRILL interconnect is not explicitly in-charter for L2VPN it would
>>>> appear to be implicitly in-charter - our decision to standardise
>>>> solutions for PBB VPLS and PBB E-VPN probably sets a precedent (since
>>>> PBB is also unmentioned by the charter).
>>>>>> 
>>>>>> On this basis we would like to ask the WG the following three
>>>> questions:
>>>>>> 
>>>>>> 1) is the WG happy for us to pursue interconnection of TRILL islands
>>>> over L2VPN?
>>>>>> 
>>>>>> 2) is the WG happy for the authors to split the TRILL section of the
>>>> PBB E-VPN draft into a separate draft?
>>>>>> 
>>>>>> 3) is the WG happy for the TRILL draft to be a WG doc?
>>>>>> 
>>>>>> Please respond by May 22nd if you are unhappy with any of these 3
>>>>>> actions (I'll take no response to mean the WG is in agreement with us
>>>>>> proceeding...)
>>>>>> 
>>>>>> Giles
>>>>>> 
>>>>>> 
>>