Some benchmarks tests – no dedicated server needed!

We have been performing a number of bench mark tests, in order to give 3CX users an idea what sort of memory and processing power is required by a 3CX installation.

First off – in many cases a dedicated server is NOT required. Modern servers have ample processing power to handle 3CX and other server applications on one box. This is also a huge advantage compared to an appliance or a linux based systems – use an existing server and save on hardware costs, energy and admin time!

To prove this point, we loaded up a modest machine with Small Business Server 2003 R2. Machine specs: Intel Core 2 Duo CPU, E 4500 @ 2.20 GHz, 4 Gb of RAM, 50 Gb Hard disk SATA, 100Mbps Network connection.

Next we loaded up IIS, Exchange Server and Active Directory and run an instance of Exchange Load simulator to simulate 25 users making heavy use of Exchange (Sending of mail, scheduling meetings, checking inbox etc)

We then installed 3CX Phone System Configuration V7.1.6591 on the same machine, selecting Cassini as the web server. We selected Cassini because it keeps 3CX independent of IIS updates and any configuration changes of OWA, Sharepoint and so on. Cassini will do fine on this environment.

We then triggered our call simulators to create a total of simultaneous calls 16 continuously. The rate of the calls was 0.5 calls/second so that in 1 hour it processes roundabout 2000 calls. Note that 16 calls simultaneously is for an install of about 40 to 50 extensions.

During this time, total CPU usage for all 3CX services  was less then 15% CPU.

Total Memory usage for all 3CX services was about 300 megabytes (With 100 megabytes taken up by Cassini)

At the same time Exchange was not using more then 10%. The machine rarely went above 30% processor usage in total, on average it was much lower.

Clearly a home run for 3CX & Windows in the game against appliances and Linux! :-)