That's probably because the TCP sockets cannot be persisted. I guess TCPNet doesn't take that into account and thus breaks on trying to restore a connection to a socket that doesn't actually exist anymore

sorry if i don't sound all that smart as i'm still new to these things, but i have been able to shutdown the luarocks server and bring it back up while my minecraft server was still going, and my scripts would reconnect after they realized that they weren't connected anymore. i don't know if that is the same as what you are describing, but i really wouldn't know the difference between the luarocks server shutting down, or the minecraft server shutting down, and why that would cause my OC computers to freeze for one thing, and not the other. this is how i tested what i wrote to make sure that it would still reconnect after it disconnected from luarocks.

guessing that the luarocks server is the external one, they will still connect because the java connection object they're using still exists. upon restarting minecraft that java socket can't be persisted so any operations on it fail.

@fnuecke might be able to make it re-create socket objects on restart but i'm not sure how easy it would be to do

Hmm, the to literally persist userdata part should be impossible, so that's definitely at least partially a bug in OC. I'll look into it.

Could you please try setting verbosePersistenceErrors to true in the config and see how the message in the log looks then?

http://pastebin.com/MirJpxRZ

here you go sir, just as you requested. that is for the 4 computers that are all set up to reconnect to TCPnet. it only shows the error when you right click on the server rack/computer that is having the issue.

Thanks, that's a good lead for a start. Looks like some upvalues aren't properly wrapped when the computers are being saved while they're in the middle of a synchronized call. Please give dev build 480 a try and see if that fixes it.

k i'll throw this on my test server and mess with it. thanks again for the quick response!

@fnuecke i threw the dev build on my test server (mirrored server all the way down to the previous 15 mins on my live server) and it looks like that fixed everything. i can restart the server with my scripts running and they will re-establish a connection after a server restarts, and my OC computers don't freeze up anymore. unless you have any other need for this to be opened, i'm going to go ahead and close this. thanks again, and keep up the outstanding work!

Awesome, thanks for confirming!

Make sure to set verbosePersistenceErrors to false again, as it's quite the performance and memory hog while saving.

yep already have. now the only thing i have noticed is that my computers won't re-establish a connection until a player gets near the computer, even with a dimensional anchor(chunk loader) loading the area. not a big deal because there is always someone near by, but if there is something else that can be done about that, it wouldn't hurt i don't think. but other then that i'm happy they just reconnect.

That's... odd. As long as the chunk is loaded the computers should run; there's no code that depends on an actual player being nearby anywhere in the mod, at least not to my knowledge (aka if that's the case it's some Minecraft idiosyncrasy I'm not aware of).

From my experience this may be caused by cauldron(assuming you use that)