The Cluster Status in RDS describes the health of an Aurora cluster. Typically, the cluster status should always be available. When the cluster status changes, Blue Matador creates an alert. The cluster status values are not fully documented by AWS, and in most cases investigating the instances in the cluster will reveal any issues.

Possible Issues

  • Failover: When a cluster is in the process of failing over, its status will change to failing-over. This can be caused by a user-initiated failover, or an automated failover.
  • Instance failure: When an instance fails, it could put the cluster in an unhealthy state depending on the type of failure. The easiest way to investigate this is to check the status of each instance in the cluster, and look at the RDS Events list to see if any failures recently occurred.