To use PerformanceGuard you need some agents to collect performance data on computers. You also need a server-side installation to process the performance data that agents collect.Most people install the server side first, and then they install agents, but you can also install agents first if you want. The important thing to know is that you need to install both.

The server-side installation can take some time, because it involves several components and some database setup. The agent installation is easy and usually takes less than a minute. If required, you can deploy agents to multiple computers via a software distribution tool.

Before you install, let's take a quick look at the main components of PerformanceGuard. That way, you'll know what you're going to install:

• Agents collect performance data on computers, so you must install an agent on each computer that you want to monitor. At regular intervals each agent sends performance data to the PerformanceGuard frontend server.
• The frontend server listens for data from agents. It never contacts the agents; it's always the agents that send performance data to the frontend server. The frontend server stores the agents' performance data in a database. Performance data from individual agents is labeled with individual IDs, so PerformanceGuard knows where each piece of data comes from. The performance data that the frontend server stores is very detailed. The frontend server then sends a copy of the performance data to the PerformanceGuard backend server. Big systems can have more than one frontend server that each sends performance data to the backend server.
• The backend server stores the performance data in its database in less detailed resolutions that take up less space and can be stored for a longer time. The backend server also has a built-in web server.
• When you want to view performance information, you open the PerformanceGuard web interface in a browser. The web interface connects to the backend server's web server. When you use the web interface, it seamlessly gets detailed performance data from the frontend server and less detailed performance data—such as data about trends—from the backend server.

PerformanceGuard by default uses these ports