Help_Topics

Set Performance Objectives / SLA

When monitoring website or application uptime and performance, it is often helpful to be able to measure current performance against your company's Service Level Agreement (SLA) or performance objectives. You can enter these defined objectives into your GlobalWatch monitoring services to quickly and easily determine whether you are meeting them.

To Set Performance Objectives

  1. From the Dashboard tab, click one of the service's [Set]  links, or click the name of the service, open the Settings tab, and select SLA/Objectives. The Performance Objective/SLA window appears.
  2. Enter a value in the Threshold box.
  3. Notification Choice specifies the type of notification:
    1. Alert: Alerts are sent to diagnostic/warning contact and alert contacts (per escalation levels).
    2. Warn: Alerts are sent to the diagnostic/warning contact only.
    3. Ignore: No alerts will be sent.
  4. In the Alert Using list, select if you want to determing your performance objectives using only the Baseline Agents (the default), or Baseline and Additional Agents (if you are using additional agents).
  5. The Only alert when the running average for the past x minutes exceeds my SLA threshold checkbox controls the violation mechanism used to calculate an SLA violation:
    1. If the checkbox is cleared (the default), the Individual Violation Mechanism alerts you when the load time exceeds the SLA load time objective from x consecutive agents (where x is the number of strikes specified in the Strikes Before Error setting (configured in the Settings sub-tab Monitoring Settings option).
    2.  If the checkbox is selected, the Running Average Mechanism alerts you when the load time average of samples within x minutes exceeds the threshold.
  6. Click Update Settings to save your changes.
  7. Open the Dashboard tab and verify that the performance objectives appear for the service.

Once performance objectives have been set, you can modify them from the Settings tab for your service. From the Service Settings menu, select the SLA/Objectives menu option to view and update the performance objectives.

Alerting for an SLA Violation

An SLA violation occurs when:

  • the daily uptime average is not met
  • a page's load time exceeds the SLA load time objective from X consecutive agents
  • a sample's load time exceeds the SLA load time objective from X consecutive agents

(X is the number of strikes specified in the Settings tab > Monitoring Settings option > Strikes Before Error setting.) The agents can be consecutive baseline agents only, or consecutive baseline and additional agents (configure agents in the service's Settings tab > Monitoring Agents option).

The levels of notification are:

  1. Alert Alerts are sent to diagnostic/warning contact and alert contacts (per escalation levels).
  2. Warn: Alerts are sent to the diagnostic/warning contact only.
  3. Ignore: No alerts will be sent.

Note

  • This setting only configures alerting and does not affect reports or logs.
  • The log file only records the final SLA violation that triggered the alert. Previous strikes for an SLA violation are executed,  but not recorded in the log file.
  • This type of alert is continuous, and does not stop after a certain number of alerts, even if the Alerting Options (in the Settings tab's Alert and Notification Settings) are set to stop after X alerts.

See also

Interpret_the_Dashboard_statistics