Amanda node managment

For a while we've been managing Amanda's multiple Lavalink nodes manually by issuing eval commands. This is annoying because we have two shards, which means that we have to make all changes twice, copypasting commands into two different guilds to access the individual shards.

I've finally put together a web frontend to make it easy for us to configure nodes. We can enable or disable each node on its own on all shards by pressing one button. A node being disabled means that newly created queues won't be assigned to it and track lookups won't be done on it. Once all existing queues end on a disabled node, we can press a button to disconnect the websocket and then shut down the Lavalink process.

This should make it much easier to manage bringing the nodes in and out as they're blocked and unblocked over time. As mentioned, this used to be annoying, and we'd have to intervene every couple of days because something would get blocked.

Related to this, I have a new creatively named project that, well actually, just click that link and read the README to find out what it's for. This should be lighter and easier to run on other servers to provide the Invidious API without the high resource usage and bugs that come with actual Invidious.

I also watched a large quantity of lecture recordings.

— Cadence

← Previous: Bibliogram updater plansNext: Byte length discoveries →