Server monitoring can refer to a range of different tools and approaches for monitoring the performance of web and application servers. At its most basic, server monitoring means collecting machine- and infrastructure-level metrics like Disk I/O, throughput, CPU utilization, and memory usage.
These metrics help operations teams get a pulse on their infrastructure, but they fall short when it comes to finding, isolating and troubleshooting problems that originate in the application code. The AppDynamics platform provides all the metrics of server monitoring tools in addition to the troubleshooting capabilities of an application performance management solution.
Server Monitoring Tools Can't Do Automatic Application Discovery
AppDynamics automatically detects, instruments, and maps your application topology, showing you where the latency exists between your different application tiers.
Server Monitoring Tools Lack Business Transaction Context
Another advantage of application monitoring over server monitoring is its ability to understand Business Transactions. With Business Transaction context, it's much easier to prioritise which performance bottlenecks are critical based on which transactions they affect.
Agentless Server Monitors Can't Do Call Stack Traces
One of the biggest limitations of server monitoring software is that they can't offer code-level insight into performance bottlenecks. Because AppDynamics sits inside the application container, it can deliver class and method-level detail around performance bottlenecks, making it easier than ever to identify and fix problems.
Server Monitoring Tools Use Basic Alerting Thresholds
Many server monitoring tools have alerting capabilities, but your alerts are only as good as the as the thresholds they're triggered on. When you set static alerts they are usually set either to high or too low. This causes either too few or too many alerts that eventually get discarded because they are inaccurate. With AppDynamics server visibility and server monitoring, you get alerts on application, Business Transaction and infrastructure metrics, based on algorithms that calculate deviation from normal behavior so you only get alerts when they really matter.