RE: : Re: Layer 2 MPLS VPN

I personally don't agree with your thoughts.  BGP - yes it scales well, but
for signalling??  So a signalling packet hits the RR which then reflects
that signal to each end point??  That's slightly silly in my mind and I
believe LDP will actually win in the long wrong.  Also I believe that the
RFC/IETF bods are nearly edging towards LDP as the signalling protocol for
vpls and such and in that instance the companies who have been arguing over
BGP being better than LDP will have to rethink their strategy.

> -----Original Message-----
> From: Harpreet Singh [mailto:harpreet.singh <at> vsnl.co.in] 
> Sent: 30 December 2004 04:59
> To: 'mohd irwan'; mpls-ops <at> mplsrc.com
> Subject: RE: [MPLS-OPS]: Re: Layer 2 MPLS VPN 
> 
> 
> There are a few SPs in Asia which are offering L2 VPNs. From 
> a signaling perspective, however I feel that BGP is much more 
> scalable and flexible.
> 
> With LDP, you would need a lot of targeted LDP sessions which 
> would not scale. You get into the conventional N square 
> problem. With BGP, you just have to have a MP BGP session 
> with a RR, which you need for a L3 VPN also. Having MPBGP for 
> signaling means using the same signaling for a host of 
> services L3/L2 which increases scalability.
> 
> Further, the auto discovery features make MPBGP signaling a 
> better choice for provisioning.
> 
> Regards

RE: : Re: Layer 2 MPLS VPN

The day you have to move a couple of hundred customers from one PE to 
another PE, I'm quite sure that you (and your customers) wont find the 
idea of using BGP for signaling and auto-discovery silly, as opposed to 
manually reconfiguring a couple of hundred l2vpns times the number of 
PE's those l2vpns span.  About RR'r, nothing stops you from meshing your 
PE's to avoid the RR - not that is makes much sence, but you can certainly 
do it.

=) Per

On Sat, 1 Jan 2005, McCallum, Robert wrote:
> I personally don't agree with your thoughts.  BGP - yes it scales well, but
> for signalling??  So a signalling packet hits the RR which then reflects
> that signal to each end point??  That's slightly silly in my mind and I
> believe LDP will actually win in the long wrong.  Also I believe that the
> RFC/IETF bods are nearly edging towards LDP as the signalling protocol for
> vpls and such and in that instance the companies who have been arguing over
> BGP being better than LDP will have to rethink their strategy.
>
>
>
>> -----Original Message-----
>> From: Harpreet Singh [mailto:harpreet.singh <at> vsnl.co.in]
>> Sent: 30 December 2004 04:59
>> To: 'mohd irwan'; mpls-ops <at> mplsrc.com
>> Subject: RE: [MPLS-OPS]: Re: Layer 2 MPLS VPN
>>
>>
>> There are a few SPs in Asia which are offering L2 VPNs. From
>> a signaling perspective, however I feel that BGP is much more

Re: : Re: Layer 2 MPLS VPN

 > About RR'r, nothing stops you from meshing your
 > PE's to avoid the RR - not that is makes much sence, but you can
 > certainly do it.

Well actually it makes some sense to mesh BGP peers ;) Of course I would 
not recommend to do it by hand but with some automation tool.

> BGP - yes it scales well, but
> for signalling??  So a signalling packet hits the RR which then reflects
> that signal to each end point??  That's slightly silly in my mind

Isn't it the case that each "end point" in VPLS will require such a 
signaling information as topology is usually full mesh ? Isn't it also 
the case that for over 6 years we are sending from the same RR all 
L3VPNs routes to all client PEs just for them to drop it on the floor if 
not needed ? And yes although the company I work for does not share this 
opinion I personally like this label block hack Kireeti came up with. 
It's just few extra bytes on top of already required BGP based auto 
discovery which everybody seems to agree to support (except few Radius, 
AAA, DNS, LDAP based proposals :).

 > Also I believe that the
> RFC/IETF bods are nearly edging towards LDP as the signalling protocol for
> vpls and such and in that instance the companies who have been arguing over
> BGP being better than LDP will have to rethink their strategy.

I am not sure if much thinking is required :) IMHO they just need to 
support LDP for signaling as well and let the customer choose.

RE: : Re: Layer 2 MPLS VPN

: question on site identifier in L3VPNs

Hi,

Just seeking a couple of clarifications,

a. Is it possible to set the "site identifier/SOO" attribute on customer
routes/prefixes when they are injected into MPBGP via an IGP at the PE?
b. Is it not possible, that incase a network administrator wishes to use
2547 and run IGP with the Providers, all he needs to remember is his "SOO"
so that the any other provider he multihomes with, can use the same value?
c. is the purpose of SOO not the same as AS number in vanilla BGP? However,
the customer need not run BGP with the provider in this case.

-thnx
Alok

: Test 2

: Mailing List Change

RE: : Re: Layer 2 MPLS VPN

Hello,

Some questions:
> Some thoughts on this thread -
>  
> LDP signaling provides more flexibility for VPLS
> because it allows different characteristic setting
> per PW such as: 
> • 
> –1. QoS setting – e.g., different PWs can have
> different reserved BW and if needed the BW can be
> adjusted dynamically (based on time of day). Also
> congestion management requires feedback for a PtP
> link 
> 

LDP? CRLDP? or RSVP-TE? there is no qos term in LDP I
have seen.

I still think that RSVP-TE is the best label
distribution protocol and BGP works best for
discovery, but there is no UNI spefication of RSVP-TE
in MPLS world. Or atleast it does not come till my PC
yet.

the property of "signalling a path when needed with
the kind of constraints as desired" is nice about
RSVP-TE.

Though if packet "no call setup switching" is what it
will be, I still do not understand what was wrong if
the labels were unique end to end :)

> 2. Sequencing: Sequencing is a PtP operation in
> nature. Sequence numbers among different PtP can
> have different “start” values. Also re-synching of
> sequence numbers are PtP operation. Furthermore,
> sequencing can be turned on/off on a per PW basis
> and allowing the operator finer control over it. 
> 

sequencing of what?

>  3. OAM: It is important to be able to check the
> health status of each PW separately because one PW
> may affect the status of the whole set (Emulated
> VLAN) 

In the implementation of LDP I have seen for L2VPNs,
LDP just carries the end point identifiers, more like
map this vlan to this label, map this  port to this
"group" and each PE then assigns the group a label.
RSVP-TE is still the preferred way to setup LSPs to
the PE loopbacks.

> 
> 4. Data-plane failure: there can be scenarios where
> an operator may want to reset a PW (bring down and
> up). In such cases, a new label is desired to be
> assigned (in case there is an issue with the label
> itself). This is inherently done with LDP protocol;
> whereas with BGP protocol and label-block scheme the
> same label gets assigned over and over again.

I dont know about the "label block" schemes, I never
did really understand it. could someone clarify it?

All you need are end point identifiers? is that right
or wrong? and all you need are ways to map end point
identifiers to some signalling method, RSVP-TE or LDP?
is that correct or incorrect? the rest just seems to
me as beating around the bush.

Why a PW/LSP cannot be setup between point A and point
B, where A and B are end hosts, using just 1 Label/tag
per hop is beyond my comprehension, considering people
hardly even use stateless/connectionless protcols any
more.

<nerdtalk>
But seems everyone wants to do something new now that
computers have done printf ("hello world \n");
</nerdtalk>

	
	
		
___________________________________________________________ 
ALL-NEW Yahoo! Messenger - all new features - even more fun! http://uk.messenger.yahoo.com

: mpls industrial status..

Hi all ,

  Can somebody plz mail me some link or reference for  state of art of mpls 
and diffserv over mpls in the industry.. ? how are companies (ISPs) using 
it. I had been into reseach in MPLS for sometime.. but unable to contact 
someone who could give me industrial or actual implementation issues and 
perspective in mpls.

  plz help and thanks.

Sudeep

On Wed, 29 Dec 2004, Mirla, Pavan Kumar wrote:

> For more information on L2 MPLS VPN's you might want to refer to
> www.vpls.org
>
>
> Thanks,
> Pavan
>
>
>
> -----Original Message-----
> From: Harpreet Singh [mailto:harpreet.singh <at> vsnl.co.in]
> Sent: Thursday, December 30, 2004 10:29 AM
> To: 'mohd irwan'; mpls-ops <at> mplsrc.com
> Subject: RE: [MPLS-OPS]: Re: Layer 2 MPLS VPN
>
> There are a few SPs in Asia which are offering L2 VPNs. From a signaling
> perspective, however I feel that BGP is much more scalable and flexible.
>
> With LDP, you would need a lot of targeted LDP sessions which would not
> scale. You get into the conventional N square problem. With BGP, you
> just
> have to have a MP BGP session with a RR, which you need for a L3 VPN
> also.
> Having MPBGP for signaling means using the same signaling for a host of
> services L3/L2 which increases scalability.
>
> Further, the auto discovery features make MPBGP signaling a better
> choice
> for provisioning.
>
> Regards
> Harpreet Singh
>
> -----Original Message-----
> From: mohd irwan [mailto:mplsvpn2004 <at> yahoo.com]
> Sent: Wednesday, December 29, 2004 9:34 PM
> To: mpls-ops <at> mplsrc.com
> Subject: [MPLS-OPS]: Re: Layer 2 MPLS VPN
>
> Good Day !!
>
> I'm Irwan from Kuala Lumpur,Malaysia. I would like to
> know if there is any service provider have start the
> Layer 2 MPLS VPN services.
>
> I can't find any of service provider in ASEAN Country
> provide this services.
>
> I need this information for my Final Year Project
> regarding Layer 2 MPLS VPN.
>
> I want to know which signalling is better from SP view
> of point.
>
> Thank You
>
> Regards
>
> Mohd Irwan
> Bsc(Hons) Degree in Networking
> Mara University of Technology
>
> http://www.uitm.edu.my
>
>
>
> __________________________________
> Do you Yahoo!?
> Take Yahoo! Mail with you! Get it on your mobile phone.
> http://mobile.yahoo.com/maildemo
>
> -------
> The MPLS-OPS Mailing List
> Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> Archive: http://www.mplsrc.com/mpls-ops_archive.shtml
>
> -------
> The MPLS-OPS Mailing List
> Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> Archive: http://www.mplsrc.com/mpls-ops_archive.shtml
>
>
> -------
> The MPLS-OPS Mailing List
> Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> Archive: http://www.mplsrc.com/mpls-ops_archive.shtml
>

--

-- 

Re: : Re: Layer 2 MPLS VPN

>----- Original Message ----- 
>From: "Puddinhead Wilson" <puddinghead_wilson007 <at> yahoo.co.uk>
>To: <mpls-ops <at> mplsrc.com>
>Sent: Saturday, January 08, 2005 5:59 AM
>Subject: RE: [MPLS-OPS]: Re: Layer 2 MPLS VPN
>

>
>
>Some questions:
>> Some thoughts on this thread -
>>
>> LDP signaling provides more flexibility for VPLS
>> because it allows different characteristic setting
>> per PW such as:
>> .
>> -1. QoS setting - e.g., different PWs can have
>> different reserved BW and if needed the BW can be
>> adjusted dynamically (based on time of day). Also
>> congestion management requires feedback for a PtP
>> link
>>
>
>LDP? CRLDP? or RSVP-TE? there is no qos term in LDP I
>have seen.

There are two diffrent ways of signalling VPLS - with control plane served
by LDP (
draft-ietf-l2vpn-vpls-ldp-05.txt ) or BGP
(draft-ietf-l2vpn-vpls-bgp-03.txt).Both documents
describe how to distribute and set-up PWE's between PE's without many
constraints on underlying
transport technology.That's it , transport tunnels may be IP based ( GRE )
or MPLS LSP's ( LDP
or RSVP-TE). You are talking about tunnel set-up in response to post  about
signalling pros and cons.

>>  3. OAM: It is important to be able to check the
>> health status of each PW separately because one PW
>> may affect the status of the whole set (Emulated
>> VLAN)
>
>In the implementation of LDP I have seen for L2VPNs,
>LDP just carries the end point identifiers, more like
>map this vlan to this label, map this  port to this
>"group" and each PE then assigns the group a label.
>RSVP-TE is still the preferred way to setup LSPs to
>the PE loopbacks.

Again , LDP as well as BGP based VPLS can have RSVP tunnels - it is
irrelevant.What is important here
is fact that in LDP based solution works through setting up targeted
sessions and exchanging FEC elements
carrying information about VC labels.This procedure is inherently
point-to-point which could help ( as some
ppls argue ) to have better ( fine grained) control over PW ( since specific
attributes to one PW can be exchanged in p2p fashion).
In contrast , in BGP based solution PE "broadcasts" blocks of usable label
for CE link ( label offset , base and range ) from which other
PE's  chose appropriate label. As you can see , in BGP solution there is no
room for signalling PW specific attributes - information
is shared by all "interested" PE's.

>>
>> 4. Data-plane failure: there can be scenarios where
>> an operator may want to reset a PW (bring down and
>> up). In such cases, a new label is desired to be
>> assigned (in case there is an issue with the label
>> itself). This is inherently done with LDP protocol;
>> whereas with BGP protocol and label-block scheme the
>> same label gets assigned over and over again.
>
>I dont know about the "label block" schemes, I never
>did really understand it. could someone clarify it?

It refers to BGP based solution - all is explained in draft :)

best regards ,

Piotr Marecki
Netia S.A