Heap memory metrics alert configuration

The heap memory-related metrics and their alert configurations for the application project are described.

Heap memory metrics

Heap memory is a data storage space allocated by the Java Virtual Machine (JVM) to run programs and stores key data such as instances and arrays. The heap area is closely related with the instance life cycle and GC. GC is the task of releasing unnecessary instances from the heap area to secure free space in the heap area.

Through heap memory-related metrics, you can check the maximum heap memories and current heap memories for each server to check the risk level. Additionally, you can determine the deviation between the maximum value and the current value to set the heap memory appropriately. If the minimum value for heap memory usage continues to rise, you may suspect a memory leak.

Metrics Search

Home > Select Project > Sitemap > Analysis > Metrics Search

The heap memory-related metrics are described that can be seen in Metrics search.

Heap Memory

• Heap commit heap_committed

  This indicates the heap memory committed for use by the JVM.

• Initial Heap heap_init

  This indicates the initial heap memory that the JVM requests to the OS for memory control.

• Heap Max Usage heap_max

  This indicates the maximum heap memory that the JVM can use for memory control.

• Heap use heap_used

  This indicates the heap memory used by the JVM.

Non-heap Memory

• Heap commit nonheap_committed

  This indicates the initial non-heap memory that the JVM requests to the OS for memory control.

• Initial Non-heap nonheap_init

  This indicates the initial non-heap memory that the JVM requests to the OS for memory control.

• Non-heap max usage nonheap_max

  This indicates the maximum non-heap memory that the JVM can use for memory control.

• Non-heap use nonheap_used

  This indicates the non-heap memory used by the JVM.

Metrics Chart

Home > Select Project > Analysis > Metrics Chart

The heap memory-related metrics are described that can be seen in Metrics chart.

• Heap total heap_tot

  It indicates the total memory for the JVM.

• Heap use heap_use

  This indicates the heap memory being used by the JVM.

• Heap Max Usage heap_max

  This indicates the maximum memory that the JVM can allocate.

• Heap perm area usage heap_perm

  This indicates the memory usage in the Perm area.

• Heap pending finalization count heap_pending_finalization

  This indicates the approximate number of objects waiting for finalization.

Note

For more information about application metrics, see the following.

Heap memory metrics alert configuration

Home > Select Project > Alert > Select Metrics tab

To use the heap memory-related alerts, configure the metrics alerts. Metrics alert notifications allow for more specific and complex alert settings rather than basic alerts. For more information, see the following. The following example explains how to configure the metrics alerts based on the heap memory usage (heap_used).

1. In Alert > Event Setting, select the Metrics tab.

2. Select + Add Alert Policy on the upper right of the screen.

3. In the area on the Metrics event window, configure settings for Event name, Category, Level, Message, Event rule, and the like.

   • Enter the Event name.

     Heap Memory

   • Level indicates the alert level of the occurred event. Set the level. The default value is Critical.

   • Select JavaMemory (java_memory) in the Category field.

   • Enter the event message. In case of the example message, when the heap memory usage exceeds the set threshold, it is sent along with the time of the occurred event.

     ${heap_used} exceeded ${metricThreshold} at ${time}.

   • Enter the fields, operators, and thresholds to set the event rule. An example of conditions that trigger an alert when the heap usage exceeds 50,000,000 bytes is as follows:

     Field heap_used, operator >, threshold 50000000

4. In the area, set the detailed requirements for event reception, and then test the event alert rules .

   

   • Notification setting

     • Number of event allows you to receive alerts when the event has occurred n' times in the last n period.

       The alert is sent when the event occurs 3 times in the last 1 minute.

     • Event pause prevents the same alert from occurring for the set time after the alert occurred.

   • The Testing event rules option allows you to see how many alerts were triggered when setting alerts for the event during the selected period.

     During the last 60 minutes, the total number of alerts whose heap memory usage (heap_used) of more than 3 times in a row exceeded 50,000,000 bytes within 1 minute is 66.

5. Select Save at the bottom of the Metrics event window to end the alert configuration.

Heap memory usage checking

Home > Select Project > Analysis > Metrics Chart

To check the heap memory usage (heap_used) in detail for the selected period, go to Metrics chart.

1. Through the time selector at the top, select a time to view.

2. From the metrics list on the left, select JavaMemory (java_memory) for the Category.

3. Under the JavaMemory category, select heap_used. In the chart area, you can see the heap memory usage chart as shown in the example.

4. Click More on the upper right of the heap_used widget, and then select Detail in the option menu. Then each metrics trend can be viewed for the monitoring targets.

   

5. Select Tag filter, then you can see individual charts by selecting values for agent, agent group, agent host, and the like.