Properties

alarm_actions

  • A list of URLs (webhooks) to invoke when state transitions to alarm.
  • Optional property.
  • Type: List

comparison_operator

  • Operator used to compare specified statistic with threshold.
  • Optional property.
  • Allowed values: ge, gt, eq, ne, lt, le
  • Type: String

description

  • Description for the alarm.
  • Optional property.
  • Type: String

enabled

  • True if alarm evaluation/actioning is enabled.
  • Optional property, defaults to "true".
  • Type: Boolean

evaluation_periods

  • Number of periods to evaluate over.
  • Optional property, defaults to 1.
  • Type: Integer

insufficient_data_actions

  • A list of URLs (webhooks) to invoke when state transitions to insufficient-data.
  • Optional property.
  • Type: List

matching_metadata

  • Meter should match this resource metadata (key=value) additionally to the meter_name.

    When monitoring the CPU usage rate of instances in the AutoScalingGroup, metadata.user_metadata.groupname is specified as the key, and the AutoScalingGroup resource is specified for the value.

    When this item is used for automatic recovery of an abnormal instance detected using the health check of Load Balancer, resource_id is specified as the key, and the name of Load Balancer is specified for the value.

  • Optional property, defaults to "{}".
  • Type: Map

meter_name

  • Meter name watched by the alarm.

    When monitoring the CPU usage rate of an instance, fcx.compute.cpu_util is specified.

    When this item is used for automatic recovery of an abnormal instance detected using the health check of Load Balancer, fcx.loadbalancing.instance.unhealthy is specified.

  • Required property.
  • Type: String

ok_actions

  • A list of URLs (webhooks) to invoke when state transitions to ok.
  • Optional property
  • Type: List

period

  • Period (seconds) to evaluate over.
  • Optional property, defaults to 60.
  • Type: Integer

repeat_actions

  • False to trigger actions when the threshold is reached AND the alarm's state has changed. By default, actions are called each time the threshold is reached.

    Alarms that occur during the cooldown period leading up to the next scale are discarded, so it may not be possible to recover from events that have occurred.

    Specify "true" to periodically issue an alarm until recovery of the event(s) in question.

    When this item is used for automatic recovery of an abnormal instance detected using the health check of Load Balancer, "true" is specified.

  • Optional property, defaults to "false".
  • Type: Boolean

statistic

  • Meter statistic to evaluate.

    When this item is used for automatic recovery of an abnormal instance detected using the health check of Load Balancer, "min" is specified.

  • Optional property.
  • Allowed values: count, avg, sum, min, max
  • Type: String

threshold

  • Threshold to evaluate against.

    Notes when automatic recovery of abnormal instances detected by the health check of Load Balancer is used:

    • Specify the same value as the ScalingAdjustment value specified in the policy settings.
    • When a value of 2 or higher has been set, automatic recovery will not take place until at least that number of instances are in an error state.
  • Required property.
  • Type: Number