@AverageGoob@lemmy.world Consider running something on it to monitor temps/memory/etc over time and push them to a separate device as it does so - that way you can measure things leading up to a crash, and know when it crashes (due to it suddenly stopping its reporting)