On 2020-08-21, 5:57 AM, "tom petch" <[email protected]> wrote:
From: Reshad Rahman (rrahman) <[email protected]>
Sent: 20 August 2020 18:42
I had noticed the lsps vs lsps-state, mentioned it at last BFD WG meeting
and have been in touch with the teas-yang authors.
I hadn't noticed that mpls:enabled had been removed. I'll have to go
through all MPLS-related items in the BFD yang.
<tp>
Yes please; enabled is still there but in a different place; I have not
gone back to mpls-base-yang-03 to see if the semantics are the same.
Note my third rather indecisive comment that mpls-base-yang now models mpls
in a different way, splitting IP routes with some MPLS, from MPLS only routes,
with no IP, as described in mpls-base-yang-15; I have not got my head around
this and do not know how it fits with bfd but suspect that it needs some
thinking about.
<RR> I don't think it has an impact because BFD uses an IP-prefix as MPLS-FEC.
But I will take a look.
I read the meeting minutes but your significant (for me) contribution
appears to have passed the minute taker by:-)
<RR> It was mentioned in the chairs slides though.
Regards,
Reshad.
Tom Petch
Regards,
Reshad.
On 2020-08-20, 12:33 PM, "t petch" <[email protected]> wrote:
Yes bfd-yang. Sometimes I would like to be wrong.
When I look at this I-D, I see that it references
/rt:routing/mpls:mpls/mpls:interface/mpls:config/mpls:enabled
In 2018, the MPLS WG removed that /config from the mpls-base-yang so
this would seem to be no longer valid. What needs changing to rectify
this I have not explored.
The I-D has
augment "/te:te/te:lsps-state/te:lsp"
which I no longer see in draft-ietf-teas-yang-te - the -state has gone.
Again, I have not explored the ramifications of this.
MPLS WG has a new base-yang out this week which differentiates between
an IP route with a MPLS next hop and a MPLS route with no IP, the latter
forming a new, mpls Address Family. I would think that the latter is
not catered for by BFD but it would be nice to be wrong
Tom Petch
