What is the maintenance window for Amazon Relational Database Service (Amazon RDS) and Amazon Redshift? What is the difference between cluster maintenance and instance maintenance?
The maintenance window for Amazon RDS and Amazon Redshift occurs weekly, and DB instances can receive upgrades to the operating system (OS) or to the DB engine. AWS requires at least a 30-minute window in your instance's weekly schedule to confirm that all instances have the latest patches and upgrades. During the maintenance window, tasks are performed on clusters and instances. For the security and stability of your data, maintenance can cause instances to be unavailable.
The maintenance window defines when the deployment or operation begins, but maintenance itself can take longer to complete. As a result, the time used for some operations can exceed the duration of the maintenance window.
Amazon RDS maintenance can affect all instances or just specific instances. To minimize the impact that Amazon RDS maintenance can have on your workload, see How do I minimize downtime during required Amazon RDS maintenance?
For all versions after Amazon Aurora 1.8, DB updates are performed cluster-wide, and maintenance is deployed during the cluster's maintenance window. DB engine updates to Aurora clusters can be performed across all instances or on a one-by-one basis, depending on the type of update. For more information, see Maintaining an Amazon Aurora DB cluster.
Amazon Redshift maintenance affects all instances that require updates for OS- or engine-specific updates. To minimize the impact that Amazon Redshift maintenance can have on your workload, schedule around maintenance windows.
What happens to Amazon RDS and Amazon Redshift queries that are running during a maintenance window?
How do I configure notifications for Amazon RDS or Amazon Redshift maintenance windows?
How long is the Amazon RDS maintenance window?