After thread pool usage surpasses a threshold, PingFederate can create an administrative notification, and can initiate a thread dump event. This allows administrators to respond quickly to thread exhaustion, and provides insights that can help with future troubleshooting.

  1. Go to System > Monitoring & Notification > Runtime Notifications.
  2. Select the Notification for Thread Pool Exhaustion Events check box.
    Note:

    For more information about thread pool monitoring settings, see Configuring runtime notifications.

  3. Click Save.
  4. To configure the settings for thread pool monitoring, open <pf_install>/pingfederate/server/default/data/config-store/com.pingidentity.monitoring.MonitoringService.xml.
  5. Configure the following parameters:
    Parameter Description

    MBeanName

    The name of the MBean used to monitor for thread exhaustion.

    MBeanAttribute

    The MBean attribute used to monitor for thread exhaustion.

    Operator

    The value that the MBean operator used to decide if the threshold was breached.

    Supported values are GreaterThan and LessThan

    Threshold

    The value that the MBean attribute is compared with to decide if the threshold was breached.

    The default value is 100.

    SampleIntervalSeconds

    The interval between checks.

    The default value is 15 seconds.

    Samples

    The number of consecutive samples breaching the threshold to trigger a thread exhaustion notification.

    The default value is 2.

    QuietPeriodMinutes

    The minimum amount of time before another thread exhaustion notification can be triggered.

    The default value is 15 minutes.
  6. After making your changes, save the com.pingidentity.monitoring.MonitoringService.xml file.
  7. Restart PingFederate.