Thread-Watcher is currently using ~370mb of ram per shard instance on the official instance. Spread across 23 shards at ~370mb each that works out to be around 8gb of ram usage just for the shard instances themselves. I'm unsure what we can do here as the discord.js client is already configured to cache very little.
things to consider:
- run fewer shards. Maybe we can fit more guilds into each shard so we get less overhead that comes from each shard instance?
- slim down d.js even more
- give up and get more
this is fairly urgent as we've not got enough ram to run the sveltekit dashboard / website and are at the brink of the bot slowing down / failing to work due to 100% ram usg
Thread-Watcher is currently using ~370mb of ram per shard instance on the official instance. Spread across 23 shards at ~370mb each that works out to be around 8gb of ram usage just for the shard instances themselves. I'm unsure what we can do here as the discord.js client is already configured to cache very little.
things to consider:
this is fairly urgent as we've not got enough ram to run the sveltekit dashboard / website and are at the brink of the bot slowing down / failing to work due to 100% ram usg