> On Aug 20, 2019, at 2:30 PM, Alberto Rodriguez Natal (natal) 
> <[email protected]> wrote:
> 
>       - TTL is limited to minute units. This may be overly restrictive. 
> Couldn't
>         there be some use (perhaps not common) e.g., perhaps when debugging, 
> or in
>         future versions of the protocol where seconds granularity might be 
> useful?
>         Changing these nodes later is non-backwards compatible and thus very
>         painful to do.
> 
> [AR] That's a fair comment. We used minutes in the model, however, since 
> RFC6833bis defines the TTLs in minutes. Do you think it would be reasonable 
> to leave the TTL in minutes and aligned with 6833bis?

If it were me I'd probably use seconds and mention in the description that the 
RFC granularity is minutes; however, since you don't define a range (a good 
thing) one backward compatible way to be able to specify sub-minute values in 
the future would be to augment in a seconds (or sub-seconds) node and have the 
minutes value be 0, so I think you're OK if you want to leave it as minutes.

Thanks,
Chris.

Attachment: signature.asc
Description: Message signed with OpenPGP

_______________________________________________
lisp mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lisp

Reply via email to