Re: I-D Action: draft-ietf-bfd-yang-06.txt

[Ashesh Mishra <[e-mail]>]

>From what I understand, the echo intervals have an inverse relationship with their control interval counterparts. Faster echo allows for slower control frame rate. So they are necessarily different values. That said, having never implemented echo mode, I can't comment how they should get grouped in the model.  

Ashesh

On Jul 31, 2017, at 4:56 PM, Jeffrey Haas <[e-mail]> wrote:

Mahesh,

> On Jul 31, 2017, at 4:26 PM, Mahesh Jethanandani <[e-mail]> wrote:
> 
> Jeff,
> 
>> On Jul 31, 2017, at 10:05 AM, Jeffrey Haas <[e-mail]> wrote:
>> 
>> On Fri, Jul 28, 2017 at 03:58:06PM -0700, Mahesh Jethanandani wrote:
>>> The changes are done and pushed to GitHub. Use the grouping client-cfg-parms.
>> 
>> Question: For implementations that use Echo mode, is that something the
>> protocol client configuration impacts or is it chosen by the system
>> automatically?
> 
> I am not sure I understand. 
> 
> For Echo mode, there is a separate grouping called echo-cfg-parms that allows clients to specify the desired-min-echo-tx-interval and required-min-echo-rx-interval. 
> 
> That is somehow different from desired-min-tx-interval and required-min-rx-interval that one can specify for a normal BFD session. For the life of me, I do not understand why the two need to be different :-(. I must be getting amnesia. Maybe one of the co-authors will jog my memory.

The protocol permits those values to be distinct from the async ones.  They should be in a separate set of params.

My point, unless my very quick glance at the module mislead me, is that you can't configure to use echo mode - if it's supported - in the grouping imported by the IGPs.

My implementation doesn't support echo, so I don't have any opinion about where configuration of those belongs or not.

-- Jeff



> 
> Cheers.
> 
>> 
>> -- jeff
> 
> Mahesh Jethanandani
> [e-mail]
> 
>