Hi
I’m trying to get updates from my clients across to the server 4 times a second, and updates from the server to the clients at 1 a second (the max timer resolution atm). The messages from the client to the server are about 44 bytes. SGS is obviously doing some queuing of the messages and here is why I think that.
I ran 2 clients for the first time the other night. I ran up SGS on another box and saw both clients being updated. Then the updates started to get delayed, then it got worse. I closed one client and it took 60 seconds before the SGS logs spat out the ‘client disconnected’ message, in the mean time it carried on with message recieved messages, and the remaining working client continued to get updates for the position of the client that had quit. These were updates sent before it closed, but the effect hadn’t been seen in either client, and here they are, now 60 seconds behind when they were sent. Then 60 seconds after I closed the client, pop, it goes from the other’s view.
I’ve tried it a number of times, and it seems that the longer I run 2 clients for, the further behind the server gets. The implication is that even if I run at 1 tick per second (which seems to solve it) that 8 clients would still kill the server.
Any ideas?
Endolf