[09:11] <daftykins> mornin' all
[09:31] <zxm-pi> shh, it's a monday out there
[09:36] <daftykins> i'm glad i remembered the fella with an iMac was bringing it over for 9am to sort his AV out, hah
[09:41] <zxm-pi> greeting them in your tux pajama wouldn't be professional enough? :-P
[09:41] <daftykins> i often wonder!
[09:42] <zxm-pi> your os/2 underwear would be wasted on them :-P
[09:46] <daftykins> LGR's Windows 3.x socks want in on this outfit
[09:50] <zxm-pi> must be wearing a bit thin by now :-)
[09:54] <daftykins> i've discovered something neat, you know these IoT cameras that are open to the world and so on? some guys managed to reverse engineer them and load on some alternative firmware that lets you do fun stuff like LAN streaming via RTSP \o/
[09:55] <daftykins> i've decided to snag one for fun, so perhaps i will have it in time to livestream the cat's birthday on Sunday xD
[14:14] <Gargoyle> Hi all. I've got an intermittent DNS lookup failure on an AWS instance. So I've tried updating resolve config to make /etc/systemd/resolved.conf look like https://pastebin.ubuntu.com/p/KpVwmddsm9/ however, I'm still getting lookup failures. A long running service is talking to two or three hosts all day long and 90% of the time lookups are working, but "randomly" they'll fail.
[14:16] <Gargoyle> Assuming AWS, Google and Cloudflare are all not having a blip at the same time, any thoughts on how I can track this down further and/or mitigate it with more caching/retries ?
[14:19] <daftykins> space separation looks so wrong to me, is that really an accepted format o0
[14:21] <Gargoyle> Yup. (https://www.freedesktop.org/software/systemd/man/resolved.conf.html)
[14:22] <daftykins> that param being Fallback suggests antics at a higher level, so do tools like nslookup and dig occasionally muck up from there too? never touched anything AWS
[14:23] <Gargoyle> not that I can tell. Even after catching one in the logs pretty quickly and immediately ssh'ing in and running dig, it resolves fine! :-(
[14:24] <Gargoyle> There's been some discussion on AWS groups of people seemingly having this same issue with some older instance types - supposedly resolved on newer ones.
[14:24] <Gargoyle> I might have to take the AWS DNS out of the loop. Will lose local internal name resolving, but it's just too unstable at the mo.
[14:34] <Gargoyle> Ahh. I wonder if the fallback doesn't cascade down to the interface level. I just tried changing the global and ended up with this: https://pastebin.ubuntu.com/p/tkg7QWZYyy/