Connection Frequently Asked Questions

Is there a way to limit the number of active connections to a Remoting SDK Server?

The Remoting SDK SessionManagers have a MaxSessions property that does exactly this.

Alternatively you could use a Pooled Class Factory with a pool size of the number of requests you want to simultaneously serve, and an option to Fail if pool limit exceeds. This way, if you have a pool size of 25, and the 26th simultaneous call is made, it will be rejected (i.e. fail with an exception on the client side). You can catch this exception and display a friendly message to the user or even let the application handle later retries automatically.

Is there any limit to the amount of concurrent connections when using the Remoting SDK?

No, the Remoting SDK itself does not impose any limit on the connections. However, other factors might influence scalability, such as the network stack, hardware capabilities, etc.

See - What do I need to do handle 100+ users?

What causes 'Socket Error #10048 Address already in use'?

It could be that you are not using KeepAlive and running out of local sockets.

Basically, every time a connection is closed, the IP stack puts it in a wait state for a minute or so. Thus, if you open/close many sockets quickly, you will eventually run out of sockets - an IP Stack limitation, not ours.