l2vpn@ietf.org
[Top] [All Lists]

Re: Two issues regarding BGP VPLS

Subject: Re: Two issues regarding BGP VPLS
From: Kireeti Kompella
Date: Wed, 5 Jan 2005 09:44:02 -0800 PST
Hi Huke,

On Mon, 3 Jan 2005, Huke Hu Chun Zhe wrote:

> Dear All,
> regarding draft " draft-ietf-l2vpn-vpls-bgp-02 "
> there are two issues :

Thanks for your comments.

> Issue 1:
>
> Problem description in Multi Homing

> In this scenario CE1 is connected to 2 different PE and thus is
> multi-homed. PE1 and PE2 both advertise the NLRIs, in PE3 BGP
> received 2 set of label blocks from PE1 and PE2 and PE3 in turn must
> have broadcast his NLRI to both PE1 and PE2. BGP here runs the best
> route selection algorithm and say select the label block from PE2 as
> best and thus notify this label block to VPLS VSI. A PW is thus
> formed between PE2 and PE3.
>
> In this whole time PE1 does not know that his label block has not
> been selected in the best route selection and thus unaware of the
> fact that VSI PE3 does not even know that PE1 exist. PE1 form a PW
> because he has received a reach NLRI from PE3; downloads and make
> VSI state up.

Actually, since iBGP requires (logical) full mesh connectivity in the
control plane, PE1 will see PE2's NLRI.  So, PE1 can know whether it
wins the path selection or not, and based on that, can decide whether
or not to establish PWs to other PEs.  Similarly for PE2.

Note that path selection for VPLS MUST yield the same result for all
PEs (i.e., if PE1 and PE2 are dual-homed to site 1 in VPLS foo, all
PEs in VPLS foo MUST pick the same PE to reach site 1.)  I'll clarify
this in the next version of the draft.

> Issue 2:  CSV problem

> The sending of Reach and Un Reach NLRI is based only on the
> availability of the AC. The availability of tunnel is not taken into
> consideration.

It is a requirement that PEs in a VPLS must be full-meshed in the data
plane.  Suppose PE1, PE2, PE5, PE7 and PE9 are in VPLS foo.  In this
case, there must be data connectivity among all these PEs, that is,
tunnels and PWs.  If even one tunnel goes down (say between PE1 and
PE9), then either PE1 or PE9 must withdraw from the VPLS.  This can be
accomplished by having the head-end of the tunnel LSP withdraw its
NLRI when it realizes that the tunnel is down.

Hope that helps,
Kireeti.
-------

<Prev in Thread] Current Thread [Next in Thread>