Port forwarding - slower than bridged adapter?
Port forwarding - slower than bridged adapter?
I'm testing VirtualBox for use in a project where there will be very heavy load coming down on a Java web application - up to 200,000 requests an hour. We have the software running on a 'real' Debian server, and I compared this with an Ubuntu Server VM running under Windows 2003 Server, both on pretty meaty hardware. Because I have not been able so far to configure the VM with a public IP address via a bridged adapter, I've been using port forwarding instead. I'm impressed by the performance, but I'm noticing a lot more timeouts and other failures to connect to the web server on the VM than on the 'real' hardware. Maybe this is just inevitable - I hope not. I'm wondering, though, whether this might be related to port forwarding and whether using a bridged adapter (if I can get it to work) might prove faster, and am looking for any useful experience in this area. The curious thing is that the difference doesn't seem to be attributable to CPU performance in the VM, because that was registering a very low value throughout the test, suggesting that the network was the bottleneck.
[SOLVED]Port forwarding - slower than bridged adapter?
Moot point anyway now. I managed to get Bridged Adapter mode working and all connection problems have disappeared. Using the Intel T server adapter, it's now amazingly solid and fast. I love VirtualBox!