I have been trying to keep Websocket connections open and available between Pi Clients around the country and a central server with mixed results. I have been experimenting with three servers: A Pi with a fixed IP address on Port 80, The same Pi on port 51042 and a Microsoft Azure server on port 80 and 443.
I have been sending a very short command from the server to 'ping' each client every few seconds.
I have used the same B4J program on each sever, changing only two things (i) the Port and (ii) the number of seconds between 'pings'.
I have used a single Client Pi, just changing its Websocket URL string to reflect the change in Port.
It seems that running the same B4J program on the same Pi with two different Ports results in very different timeouts: On Port 80 it is about 60 seconds and on Port 51042 it is about 5 minutes.
Does this sound right?
Can I change the Timeout setting on the B4J Server for each Port number?
If yes - then will the same setting work on the Azure server too?
I have been sending a very short command from the server to 'ping' each client every few seconds.
I have used the same B4J program on each sever, changing only two things (i) the Port and (ii) the number of seconds between 'pings'.
I have used a single Client Pi, just changing its Websocket URL string to reflect the change in Port.
It seems that running the same B4J program on the same Pi with two different Ports results in very different timeouts: On Port 80 it is about 60 seconds and on Port 51042 it is about 5 minutes.
Does this sound right?
Can I change the Timeout setting on the B4J Server for each Port number?
If yes - then will the same setting work on the Azure server too?