I am running Untangle as the primary router on a few servers at our colo and I am having some issues with timeouts. I am using Blitz to profile our performance and at around 200 concurrent requests the timeouts begin to appear and get much worse as the load increases. This must be a configuration issue on my end, anyone have any ideas?
Things I know:
- During this the router shows "low" CPU usage.
- The webserver has very minimal CPU/Disk usage.
The router and webserver are 2 separate servers (I am not running VM on these) but both are identical 1U servers. They are 4x2.8GHz Intel CPUs with 6GB of ram on Supermicro boards. The router is using the 2 onboard NIC provided by the board.