The need to restart server every so often to avoid excessive ram usage bit is very interesting to me. This sounds like some issue with memory management. Not necessarily a leak, but maybe something like server keeping unnecessary references so the object cannot be dropped.
Anyway, from my experience Rust developers love debugging such kind of problems. Are Lemmy Devs aware of this issue? And do you publish server usage logs somewhere to look deeper into that?
The need to restart server every so often to avoid excessive ram usage bit is very interesting to me. This sounds like some issue with memory management. Not necessarily a leak, but maybe something like server keeping unnecessary references so the object cannot be dropped.
Anyway, from my experience Rust developers love debugging such kind of problems. Are Lemmy Devs aware of this issue? And do you publish server usage logs somewhere to look deeper into that?