Re: [babel] Restarting nodes and seqno requests

[Toke Høiland-Jørgensen <[e-mail]>]

Juliusz Chroboczek <[e-mail]> writes:

>> No, don't think so, it's just an implementation issue: Bird currently
>> won't resend the same request until it expires after two seconds. So
>> when the triggered update arrives with S'' (that is still too low), it
>> won't ask for another seqno increase. But that is a straight-forward
>> fix, and with that it converges reasonably quickly (as long as the seqno
>> is not too high I guess...)
>
> Not sure about that -- it will cause extra background traffic, which may
> have unintended consequences.  Note that seqnos are 16 bits, so converging
> to a feasible seqno takes 16384 packets on average, 32767 worst case.

Yeah, suppose that would be quite a bit of traffic...

> I'd recommend:
>
>   - add an option for picking a different router-id each time (but
>     don't make it the default, since it makes management more difficult
>     without stable router-ids);

Hmm, actually, since Bird sets its router ID from an IPv4 address, the
top 32 bits are always zero (for compatibility with the notion of router
ID that Bird uses for other protocols). My implementation of the
randomisation only randomises the top 32 bits, so there's still a stable
identifier. Maybe it would make sense to default to this behaviour?

>   - document the issue;
>   - bug the BIRD guys about implementing persistent storage; two octets is
>     all we're asking for.

Yeah, 16 bits should be enough for everyone, right? ;)

> Should I document the issue in 6126bis?  I'll definitely add a paragraph
> in applicability-statement.

Hmm, documenting it *somewhere* is needed, certainly. Maybe putting it
into 6126bis is a good idea, since I have a nagging suspicion that not
everyone will read the applicability statement before implementing it... ;)

-Toke