On Fri, Jul 29 2022 at 1:43 PM, Erik Kline
<ek.ietf@xxxxxxxxx>
wrote:
On Fri, Jul 29, 2022 at 12:48 PM Ross Finlayson <finlayson@xxxxxxxxxxx> wrote:
> On Jul 29, 2022, at 12:37 PM, Deen, Glenn (NBCUniversal) <Glenn.Deen@xxxxxxxxxx> wrote:
>
> I think that’s part of it, but it’s also getting the various sites that consume the data to update the data they are looking at.
Maybe - but why Switzerland? I don’t think we’ve ever met there?113 was in Vienna, so that is odd.
The IETF address space is officially registered to the Internet Society, in Geneva, CH.
One problem, as I understand it, is that consumers of https://noc.ietf.org/geo/google.csv might also be validating the prefix to see if it's announced anywhere. If the prefix isn't announced early enough (whenever a geo processing cycle is underway) then consumers might discard the geo updates. I wonder if the IETF prefixes are advertised from anywhere in between meetings, or only just before/as the conference network is coming up.
The address space gets announced from our AS number, at the location/hotel, starting quite a while before the meeting -- when exactly depends on when circuits get delivered and BGP turned up.
We do publish RFC8805 data, and RFC 9092 RPSL / "whois" records, and many geo-location providers consume them (including Google)
It appears that this time MaxMind didn't update for some reason - I suspect that because the address space was not used for so long during COVID they assumed it had been sold / was dormant / similar and purged the "this network is special" tags.
We (the NOC) are following up with them.
W
I'm not sure what standard processing cycles are, but I'm sure it is on the order of several weeks (perhaps only monthly).
