Servers are the target domains of your load testing. For instance, http://dmoz.loadstorm.net would be an example of a server.
In order to conduct a load test, the target must be a URL where a web site or application resides. We simply call this the “server”, and it should not be confused with a specific piece of hardware. Servers can be hardware machines running in a data center, but LoadStorm considers a server as a logical entity that will receive HTTP traffic during the load test.
The test will simulate numerous users taking certain actions on the target server. You simply give LoadStorm the URL of the web application’s page where you want the test to begin, along with the protocol and port number, and the target server is set.
Note: the server must be under your control, thus a valid target upon which LoadStorm will generate heavy volume. Otherwise our system could be used a weapon for denial of service attacks, and we of course cannot allow that. It is easy to verify the target server is yours by placing a verification code on your default page or a verification file in your root directory.
You may create as many test plans as you wish against a particular server. Also, servers can be ignored if you don’t want to actually send traffic to that domain. A good example is Google Analytics. LoadStorm will detect that your pages are calling Google Analytics through a JavaScript, and during a load test there could be many thousands of hit to the Google server as your site registers hits. Those hits are probably not desirable for counting in Analytics, so you can IGNORE the Google Analytics server that appears in your server list. This will tell LoadStorm to not initiate a post to that server.