For me, anecdotally, it's .env. It's amazing how many requests are innocently trying to find an environment on my server.
The second though is a favicon, and that one I genuinely have no idea about. The server does not host websites and there is no reason (nor way) for a browser to connect: why would nefarious people or bots or scrapers be looking for a favicon? Some kind of fingerprinting?
