| 18 Oct 2025 |
 Vladimír Čunát | So that it's still possible to switch fast. | 19:40:06 |
 hexa | default ttl is 1h | 19:41:45 |
| hexa | that's reasonably short imo | 19:41:49 |
| Vladimír Čunát | Well... the Netlify records get returned with 80s to me. | 19:50:02 |
| Vladimír Čunát | Or 120 now, say www.nixos.org. | 19:50:21 |
| Vladimír Čunát | I meant it like - if the static record turns out horrible, it's nice to be able to revert within minutes. | 19:51:16 |
| Vladimír Čunát | (and performance benefit of TTLs above several minutes seem rather low in practice, from what I've heard) | 19:51:59 |
| Vladimír Čunát | * (and performance benefit of TTLs above several minutes seem rather low in practice, from what I've heard, except maybe stuff like DNSKEYs and nameserver records) | 19:52:23 |
| Vladimír Čunát | I see it switched now. And pretty snappy clicking around the web, from here in .cz at least. | 19:56:38 |
| hexa | ideally we'd have something that we can put behind fastly | 22:30:14 |
| hexa | because fastly does give us ip addresses | 22:30:20 |
| 20 Oct 2025 |
|  @felix.schroeter:scs.ems.host changed their display name from Felix Schröter to Felix Schröter (🌄 27.10. – 09.11.). | 08:34:10 |
| @felix.schroeter:scs.ems.host left the room. | 09:44:53 |
| 21 Oct 2025 |
|  @echobc:matrix.org joined the room. | 18:10:42 |
|  mjolnir banned @echobc:matrix.org (<no reason supplied>). | 18:10:45 |
 7c6f434c | Not urgent, but would like to understand eventually.
OpaVote says elections@nixos.org is bouncing. Last time I expected to receive something specific from them to that address, it got delivered.
When someone has time, could it be checked if there was any bounce logged for an email from opavote.com domain recently (in the last 10 days I guess)?
Thanks! | 19:56:55 |
| hexa | no mention of "opavote" (case insensitive) in the postfix log of the last 14 days | 21:46:17 |
| hexa | that would be from oct 7th | 21:46:31 |
| 7c6f434c | Thanks! No clue what bounce they are complaining about then, but oh well. | 21:47:00 |
 Lun | Could it be related to https://github.com/NixOS/infra/issues/873 if that' | 21:54:09 |
| Lun | * Could it be related to https://github.com/NixOS/infra/issues/873 if that's impacting MX lookups? | 21:54:16 |
| hexa | no, that issue was only affecting ALIAS records for resolvers that were directed to the milano pop of Gandi | 23:02:38 |
| hexa | and we only had alias records on the apex, since we can't have CNAMEs there | 23:02:59 |
| 22 Oct 2025 |
| Vladimír Čunát | MX is on the apex, but ALIAS is only about address records, so I don't think MX queries were affected. | 05:19:50 |
| Vladimír Čunát | Man, Gandi support is being really unhelpful about the ALIAS issue. | 05:29:29 |
 emily | I heard Gandi has gone downhill the past years... | 14:21:07 |
| hexa | it has | 15:47:38 |
| hexa | their motto for a long time was "No bullshit" | 15:48:01 |
| hexa | dnscontrol maps an alias to a cname on a non apex name | 15:52:39 |
| hexa | sigh | 15:52:42 |
