AppNeta Performance Manager (APM) illustrates the performance of web applications from the perspective of the end user using synthetic traffic generated by one or more geographically distributed appliances. There’s a lot packed into that sentence so break it down one by one.
APM tells the story from the perspective of the end user. This is in contrast to tracking server statistics like cpu and memory usage, and either defining what is performant based on that; or inferring from that what the user is experiencing. Instead, APM defines performant simply as whether your users are happy or frustrated.
The way APM acquires this end user perspective is by monitoring from multiple, remote locations—this is geographically distributed part. From these locations AppNeta appliances are affected by WAN links, content delivery networks, load balancers, etc, just like real users are.
Now for the synthetic traffic part. Appliances run a script at regular intervals to simulate a workflow that real users might commonly go through. While humans aren’t executing these actions, our appliances generate genuine network traffic, server requests, and load. As such, you will see corresponding log entries and commensurate server load.
The advantage the kind of continuous monitoring that APM offers is that you can see how application performance changes over time—both at the macro level, as your application evolves, and at the micro level, as server and network load varies over a day or week.