Without picking through all of them, it seems like quite a few are simple errors that have just gone unnoticed for some time. A syntax error in scene or startup Lua wouldn’t impact the entire system until after a reload, so to the casual observer all may seem fine, until suddenly it’s not.
That still does not explain all of them, that is true, but given what we know about memory consumption and leaks, it’s not entirely surprising that after some random time, if memory is impacted, a sudden inability to execute Lua cleanly occurs. Many have said that a reboot of their system has resolved their issue, which for me makes some memory leak/exhaustion a high likelihood. Consider, there may be code paths rarely taken, that suddenly and just for the moment and circumstances of the day are more frequent. With big transitions going on at Vera, and a recent spate of downtime events for various subsystems, who knows what ripple effect we might see. For example, what if there’s a memory leak in the code that checks for plugin updates? If apps.mios.com or its loadshared aliases aren’t available, an error exit path is repeatedly taken as your Vera keeps retrying, but this path leaks memory or fails to return some other used resource (file handle, socket, etc.), eventually to destruction. This could even happen outside of LuaUPnP, in any of the numerous scripts and daemons that run in the system, slowly eroding some resource until LuaUPnP itself is affected.