Monitor with CloudWatch alarms - Application Auto Scaling

Monitor with CloudWatch alarms

You can create alarms to notify you when Amazon CloudWatch has detected any problems that might require your attention.

A CloudWatch alarm watches a single metric. It invokes one or more actions only when the alarm state changes, and has persisted for the period that you specify. For example, you can set an alarm that notifies you when a metric value falls to or exceeds a certain level, ensuring that you are notified before a potential problem occurs.

CloudWatch also allows you to set an alarm that notifies you when the metric is in INSUFFICIENT_DATA state. Any metric, for any AWS service, can alarm on INSUFFICIENT_DATA. This is the initial state of a new alarm, but the alarm state also changes to INSUFFICIENT_DATA if CloudWatch metrics become unavailable, or not enough data is available for the metric to determine the alarm state. For example, AWS Lambda emits the ProvisionedConcurrencyUtilization metric to CloudWatch every minute only when the Lambda function is active. If the function is inactive, this results in the alarm going to the INSUFFICIENT_DATA state while waiting for the metrics. This is normal and might not necessarily mean that there is a problem, but it could be indicative of a problem if you expected activity within a period of time but there was none.

This topic explains how to create an alarm that sends a notification when the metric is within or outside a threshold that you define, or when there is insufficient data. For more detailed information about alarms, see Using Amazon CloudWatch alarms in the Amazon CloudWatch User Guide.

To create an alarm that sends email

  1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

  2. In the navigation pane, choose Alarms, Create Alarm.

  3. Choose Select Metric.

    You are directed to a page where you can find all of your metrics. The types of metrics available to you depends on the services and features that you use. Metrics are grouped first by the service namespace, then by the various dimension combinations within each namespace.

  4. Select a metric namespace (for example, Lambda) and then a metric dimension (for example, By Function Name).

    The All metrics tab displays all metrics for the selected dimension and namespace.

  5. Select the check box next to the metric that you want to create an alarm for, and then choose Select metric.

  6. Configure the alarm as follows, and then choose Next:

    • Under Metric, select an aggregation period of 1 minute or 5 minutes. If you use one minute as an aggregation period for a metric, there will be one data point every minute. The shorter period creates a more sensitive alarm.

    • Under Conditions, configure your threshold, for example, the value that the metric must exceed before a notification is generated.

    • Under Additional configuration, for Datapoints to alarm, enter the number of data points (evaluation periods) during which the metric value must meet the threshold conditions to trigger the alarm. For example, two consecutive periods of 5 minutes would take 10 minutes to trigger the alarm.

    • For Missing data treatment, keep the default and treat missing data points as missing.

      Some metrics are reported only when there is activity occurring. This can result in a sparsely reported metric. If a metric is frequently missing data points by design, the state of the alarm is INSUFFICIENT_DATA during those periods. To force the alarm to maintain the previous ALARM or OK state to keep alerts from flapping, you could choose to ignore missing data instead.

  7. Under Notification, choose or create an SNS topic to notify when the alarm is in ALARM state, OK state, or INSUFFICIENT_DATA state. To have the alarm send multiple notifications for the same alarm state or for different alarm states, choose Add notification.

  8. When finished, choose Next.

  9. Enter a name and, optionally, a description for the alarm, and then choose Next.

  10. Choose Create alarm.

To check the state of your alarms

  1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

  2. In the navigation pane, choose Alarms to see a list of alarms.

  3. To filter alarms, use the drop-down filters next to the search field, and choose the filter option that you want to apply.

  4. To edit or delete an alarm, select the alarm and then choose Actions, Edit or Actions, Delete.