Thanks for your reply, Paolo. My remarks below.

On Jan 23, 2019, at 9:04 AM, Paolo Lucente <[email protected]> wrote:
> 
> Hi John,
> 
> [shared text] Ideally, i would like every BMP message type to have a 
> (optional) TLV section, each with a own namespace. If the idea is shared, i’d 
> look forward to see how to get there.

This is clearly an idea all right-thinking people will support and the reason 
we don't see more "+1" replies is because everyone assumes it obviously a good 
idea. :-) 

>> On 15 Dec 2018, at 00:40, John Scudder <[email protected]> wrote:
>> 
>> [ .. ]
>> 
>> One fix, I think the more expedient one, would be to request a new reason 
>> code ("TBD3"), whose format includes both the content carried in reason 2, 
>> and a VRF/Table Name. Then if you want to send the table name, you use the 
>> new reason code.
>> 
>> An alternate fix, somewhat more involved but possibly useful in the long 
>> run, would be to update RFC 7854 to permit Peer Down messages to carry 
>> additional data beyond the reason code and its own associated data, and then 
>> to define the format of that data. Probably the additional data would be 
>> TLVized. It would obviously need to have a registry created for it.
> 
> If i get you correctly, updating RFC7854 is what would get to the ideal 
> scenario above. I feel, for the reasons you explain, this is the kind of 
> change that would require a version bump in order to make it clean.

While there are ways to do it without a version bump, I agree with you that 
bumping the version keeps it clean.

> Perhaps a longer term plan, where we do also tackle the subject of TLV & 
> Route Monitor message?

As in Henk's earlier (and still not dead!) proposal?

> Although not going in the ideal direction, for shorter-term I was thinking 
> about somewhat a mix of the two solutions you propose, to work as a Charon: 
> use a new reason code (or perhaps two, one for local terminated session, one 
> for remotely terminated session) since, as you said in your follow-up email, 
> it is the more conservative and would give the most hope against what has 
> been already coded. And make this/these new reason code(s) carry "additional 
> data [that] would be TLVized. It would obviously need to have a registry 
> created for it” so not to make all too “expedient” and revolving around the 
> specifics of VRF/Table Name and draft-ietf-grow-bmp-local-rib. 

If I understand you correctly this seems fine to me. I guess the easiest way to 
be fully clear would be to rev the draft, and then we can have a look at the 
new text?

Thanks,

--John
_______________________________________________
GROW mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/grow

Reply via email to