When you call PutAnomalyDetector, the operation creates a new anomaly detector if one doesn't exist, or updates an existing one¶

Description¶

When you call put_anomaly_detector, the operation creates a new anomaly detector if one doesn't exist, or updates an existing one. Each call to this operation triggers a complete retraining of the detector, which includes querying the minimum required samples and backfilling the detector with historical data. This process occurs regardless of whether you're making a minor change like updating the evaluation interval or making more substantial modifications. The operation serves as the single method for creating, updating, and retraining anomaly detectors.

Usage¶

prometheusservice_put_anomaly_detector(workspaceId, anomalyDetectorId,
  evaluationIntervalInSeconds, missingDataAction, configuration, labels,
  clientToken)

Arguments¶

• workspaceId

  [required] The identifier of the workspace containing the anomaly detector to update.

• anomalyDetectorId

  [required] The identifier of the anomaly detector to update.

• evaluationIntervalInSeconds

  The frequency, in seconds, at which the anomaly detector evaluates metrics.

• missingDataAction

  Specifies the action to take when data is missing during evaluation.

• configuration

  [required] The algorithm configuration for the anomaly detector.

• labels

  The Amazon Managed Service for Prometheus metric labels to associate with the anomaly detector.

• clientToken

  A unique, case-sensitive identifier that you provide to ensure the idempotency of the request.

Value¶

A list with the following syntax:

list(
  anomalyDetectorId = "string",
  arn = "string",
  status = list(
    statusCode = "CREATING"|"ACTIVE"|"UPDATING"|"DELETING"|"CREATION_FAILED"|"UPDATE_FAILED"|"DELETION_FAILED",
    statusReason = "string"
  ),
  tags = list(
    "string"
  )
)

Request syntax¶

svc$put_anomaly_detector(
  workspaceId = "string",
  anomalyDetectorId = "string",
  evaluationIntervalInSeconds = 123,
  missingDataAction = list(
    markAsAnomaly = TRUE|FALSE,
    skip = TRUE|FALSE
  ),
  configuration = list(
    randomCutForest = list(
      query = "string",
      shingleSize = 123,
      sampleSize = 123,
      ignoreNearExpectedFromAbove = list(
        amount = 123.0,
        ratio = 123.0
      ),
      ignoreNearExpectedFromBelow = list(
        amount = 123.0,
        ratio = 123.0
      )
    )
  ),
  labels = list(
    "string"
  ),
  clientToken = "string"
)