Using host and process health monitoring

What is health monitoring?

A healthy system performs better. AppMon uses defined parameters to monitor health. These parameters use metrics such as CPU, memory, network, and disk. JVMs and CLRs may use metrics such as GC utilization and CPU usage.

Goal of this tutorial

To explain health monitoring for hosts and processes.

Scenario

You need an AppMon installation with several deployed agents. Use your own application, or use the easyTravel demo application to learn about various AppMon features and functions.

How does AppMon collect host information? For information, see Setting Up Infrastructure Monitoring.

Detailed steps

To calculate the host health, AppMon Agents capture CPU, memory, network, and disk metrics. To calculate process health, Java and .NET Agents capture JVM and CLR memory pool sizes and garbage collection activity.

Host Health

Host health is automatically calculated based on CPU, memory, network, and disk metrics. You can see these values in the Transaction Flow on the respective agent node. Use this to identify the impact of an unhealthy host on your business transactions.

The Transaction Flow Dashlet
The Transaction Flow Dashlet

Click Show Host Health in the Transaction Flow dashlet, or open the Host Health dashboard through the Start Center for an overview of the host metrics and the health indicators.

The Host Health Dashboard
The Host Health Dashboard

AppMon also determines the impact of an unhealthy host on your application performance and business transactions. This lets you concentrate on the application and the underlying infrastructure, if you have application performance problems. If a host runs out of disk storage, it is recommend you troubleshoot this rather than analyze application exceptions or log messages.

Process Health

AppMon Agents also collect detailed process health information that you can see in the Transaction Flow dashlet. From here, go to the Process Health dashboard for a detailed overview of CPU, memory, threads, and transaction throughput.

The Process Health Dashboard
The Process Health Dashboard

If you suspect a memory leak because of increasing heap spaces, trigger a memory dump from here directly. This is also true for high CPU or thread numbers. Simply trigger a CPU sampling session or thread dump from this dashboard.