Re: [mpls] entropy label indicator label indraft-kompella-mpls-entropy-label

[Stewart Bryant <[e-mail]>]

  On 27/07/2010 15:35, Kireeti Kompella wrote:
> Hi Stewart,
>
> On Jul 27, 2010, at 07:24 , Stewart Bryant wrote:
>
>> John
>>
>> Why do you prefer to use an ELI as opposed to using a new set of FECs with the property that they are followed by an EL?
> The ELI is a catch-all, in case there is any potential confusion that the EL may be interpreted as an app label.  Most app labels (PWs (via T-LDP or BGP), VPN labels, etc.) can easily be followed by an EL, as Shane's slides showed.  No need for new FECs.
>
> The sticky case is "naked" IP over LDP tunnels.  Do you have a suggestion that doesn't need an ELI in that case?
Hi Kireeti,

Well you could do this with a new IP FEC, but you would have to have 
this running on all P routers along the path and you would need to  PHP 
both the LSP and the entropy labels.

It's not clear which is worse here. Though I note that the core 
philosophy of MPLS is to signal in the control plane and not in the data 
plane.

Stewart