Tutorial: Creating an Amazon CloudWatch alarm for Multi-AZ DB cluster replica lag - Amazon Relational Database Service

Tutorial: Creating an Amazon CloudWatch alarm for Multi-AZ DB cluster replica lag

You can create an Amazon CloudWatch alarm that sends an Amazon SNS message when replica lag for a Multi-AZ DB cluster has exceeded a threshold. An alarm watches the ReplicaLag metric over a time period that you specify. The action is a notification sent to an Amazon SNS topic or Amazon EC2 Auto Scaling policy.

To set a CloudWatch alarm for Multi-AZ DB cluster replica lag

  1. Sign in to the AWS Management Console and open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

  2. In the navigation pane, choose Alarms, All alarms.

  3. Choose Create alarm.

  4. On the Specify metric and conditions page, choose Select metric.

  5. In the search box, enter the name of your Multi-AZ DB cluster and press Enter.

    The following image shows the Select metric page with a Multi-AZ DB cluster named rds-cluster entered.

    
                    Select metric page
  6. Choose RDS, Per-Database Metrics.

  7. In the search box, enter ReplicaLag and press Enter, then select each DB instance in the DB cluster.

    The following image shows the Select metric page with the DB instances selected for the ReplicaLag metric.

    
                    Select metric page with DB instances selected for the ReplicaLag metric

    This alarm considers the replica lag for all three of the DB instances in the Multi-AZ DB cluster. The alarm responds when any DB instance exceeds the threshold. It uses a math expression that returns the maximum value of the three metrics. Start by sorting by metric name, and then choose all three ReplicaLag metrics.

  8. From Add math, choose All functions, MAX.

    
                    The Add math setting
  9. Choose the Graphed metrics tab, and edit the details for Expression1 to MAX([m1,m2,m3]).

  10. For all three ReplicaLag metrics, change the Period to 1 minute.

  11. Clear selection from all metrics except for Expression1.

    The Select metric page should look similar to the following image.

    
                    The Select metric page with the metric selected
  12. Choose Select metric.

  13. On the Specify metric and conditions page, change the label to a meaningful name, such as ClusterReplicaLag, and enter a number of seconds in Define the threshold value. For this tutorial, enter 1200 seconds (20 minutes). You can adjust this value for your workload requirements.

    The Specify metric and conditions page should look similar to the following image.

    
                    The Specify metric and conditions page
  14. Choose Next, and the Configure actions page appears.

  15. Keep In alarm selected, choose Create new topic, and enter the topic name and a valid email address.

    
                    The Configure actions page
  16. Choose Create topic, and then choose Next.

  17. On the Add name and description page, enter the Alarm name and Alarm description, and then choose Next.

    
                    The Add name and description page
  18. Preview the alarm that you're about to create on the Preview and create page, and then choose Create alarm.