Configure a simple anomaly detector

Latest Dynatrace

Davis Anomaly Detection - new Davis Anomaly Detection allows you to create anomaly detectors, set up customized alerts, and transform metric events configuration. You can also save time and create an anomaly detector in Notebooks Notebooks while using the app.

Create a simple anomaly detector

Prerequisites

To use the latest version of Davis Anomaly Detection - new Davis Anomaly Detection, you need to have appropriate permissions. For more information, see Davis Anomaly Detection app overview.

Create or edit a simple anomaly detector

To manually create a simple Anomaly Detection configuration

  1. Go to Davis Anomaly Detection - new Davis Anomaly Detection.
  2. Select Add Anomaly Detector > Create your own Anomaly Detector to create a new anomaly detector. To edit an existing anomaly detector, select any anomaly detector from the list.
  3. In the Simple tab, expand Set scope and provide the DQL query to fetch your data.

    We recommend that you use the interval: 1m parameter to ensure proper data resolution for the analysis.

  4. Expand Define alert condition.
  5. In Select use case, choose the preferred analyzer. For details, see Analyzer type and parameters.
  6. In Set a condition > Threshold, select Suggest values to if you want Davis AI to automatically suggest a value based on the latest behavior of your data. You can also choose the desired threshold value and the Unit of your value manually.
  7. optional In Set a condition > Alert condition, select
    • Alert if metric is above to receive alerts when the value exceeds the threshold value.
    • Alert if metric is below to receive alerts when the value is below the threshold value.
  8. optional Select Preview to see a demonstration of your alert condition.
  9. Expand Add details.
  10. Set Title of your anomaly detector to any name you like.
  11. Set Event name to any name you like. The Event name will show as a title for events events generated by this anomaly detector.

    You can write { to let Davis AI suggest you placeholder names with desired value (for example, {alert_condition}). For more information, see Event template.

  12. Select Create to create a simple anomaly detector or select Save to update your configuration.

Whenever you Create or Save your anomaly detector, its configuration gets automatically validated. If the there's no errors present in your configuration, you'll be able to save or update your configuration. If there are any errors, the section will be highlighted with red and marked with Error message under the section title.

Check the Status of the new configuration shortly after creation to ensure there are no errors in the execution.

Create a simple anomaly detector in Notebooks

With Davis for Notebooks, you can preview your Anomaly Detection configuration and evaluate its effectiveness. This option takes you to Notebooks Notebooks, where you configure the query and monitoring strategy, and then back to Davis Anomaly Detection - new Davis Anomaly Detection to create an event template.

  1. Go to Davis Anomaly Detection - new Davis Anomaly Detection.
  2. Select Add Anomaly Detector > Anomaly Detector in Notebooks.
  3. Select a notebook in which you want to preview your configuration.
    This action takes you to Notebooks Notebooks.
  4. Add a new DQL or Metrics section and query the data you're interested in.

    For a DQL query, we recommend that you use the interval: 1m parameter to ensure proper data resolution for the analysis.

  5. Select Options > Davis AI.
  6. Activate the analyzer.
  7. Select the required analyzer and configure it. For details, see Anomaly detection configuration.
  8. Select Run analysis.
  9. Once you're satisfied with the result, select More actions > Open with Open with and select Davis Anomaly Detection.
    This action takes you back to Davis Anomaly Detection - new Davis Anomaly Detection.
  10. Expand Add details and set Title to any name you like.
  11. Set Event name to any name you like. The Event name will show as a title for events events generated by this anomaly detector.
  12. Select Create.

    Check the Status of the new configuration shortly after creation to ensure there are no errors in the execution.

Transform a metric event configuration

Metric events enhance anomaly detection, expanding them beyond out-of-the-box use cases into metric-based events. With the power of DQL, you can extend this reach further.

To convert a metric event to the Anomaly Detection configuration

  1. Go to Davis Anomaly Detection - new Davis Anomaly Detection.
  2. Select Add Anomaly Detector > Improve metric events with DQL.
  3. Select the required metric event and select Transform.
  4. To configure an auto-adaptive threshold and seasonal baseline, adapt the query to a 1-minute resolution.
    1. For the newly created configuration, select the transformed anomaly detector from the list.
    2. Expand the Set scope.
    3. Append the interval:1m parameter to the query.
    4. Select Save to save your changes.
  5. The converted metric event is automatically disabled, and the newly created configuration is active instead.

    Check the Status of the new configuration shortly after creation to ensure there are no errors in the execution.