Re: [Idr] Vendor Defaults (was Re: Review of draft-ietf-large-community-06.txt)

[Robert Raszuk <[e-mail]>]

Hi Brian,

> Also, I am not aware of whether attribute-sending is flagged in the
capabilities
> negotiation of an OPEN. That would, IMHO, be a better place to signal
> attribute passing vs dropping, rather than on each UPDATE. What do
> vendors think of that (or is it already done?)?

Large Community does not require capability negotiation hence its support
or not
will not be signaled in OPEN msg.

Assuming that you are proposing a new capability which would indicate if a
given
transitive attribute it to be propagated or not based on local
configuration that
while may be nice idea requires dynamic capabilities which as you very well
know is for number of years and after number of attempts to support it
nowhere.

Rgs,
R.

PS. Perhaps the right place for this would be Operational Message which
Rob, Dave
and myself proposed few years back but due to lack of support the idea died
in spite
of already being a WG supported draft ...

https://tools.ietf.org/html/draft-ietf-idr-operational-message-00





On Mon, Nov 7, 2016 at 10:09 PM, Brian Dickson <
[e-mail]> wrote:

>
>
> On Mon, Nov 7, 2016 at 12:46 PM, Robert Raszuk <[e-mail]> wrote:
>
>> Hi Brian,
>>
>> > Maybe folks from various vendors (including open source
>> implementations)
>> > could chime in on the availability of knobs for changing global
>> (default)
>>
>> Maybe we should move this thread to GROW maybe not, but since you started
>> the pool here I would add a question to indicate by any implementation that
>> if it drops an attribute (as it is not enabled to be sent over a session)
>> would it log it in syslog and if so with what error level. The worse
>> operationally are silent drops in any control plane.
>>
>>
>>
> I would definitely be in favor of including that detail.
>
> Also, I am not aware of whether attribute-sending is flagged in the
> capabilities negotiation of an OPEN. That would, IMHO, be a better place to
> signal attribute passing vs dropping, rather than on each UPDATE. What do
> vendors think of that (or is it already done?)?
>
> Brian
>