My comment was not intended as an attack, but that appears to be the way it was taken. I was trying to point out that if a server/hub (meaning it should be un-managed and continue functioning over a long period of time) requires regular rebooting due to a problem with the way it's put together (hardware, software, drivers, etc), then the solution for the developer is to fix the problem. Writing a script to reboot is somewhat akin to putting electrical tape over the CEL on your car's dash so you don't see that the car has a problem. Ignoring it may be ok for a short time, but eventually it will likely lead to other issues.
The reason tech support tells users to reboot is that they know 90% of the time they won't be able to fix the real problem, so they ask for the customer to clear the error making it "fixed". I was first line phone support for my campus network in the 90's when people who didn't know anything about computers were just showing up to school and that was in the guide for how to help. The next question was "is the network cable plugged in?"
If you just want it to work for a few hours at a time between reboots, nightly reboots will suffice. But, I've had an RPi running for months without issue (2 actually). The only reason for reboots was power issues or tinkering with configs/installing software/etc.
