RDS Multi-AZ vs read replicas: failover vs scaling reads

Try 3 questions on this

Pick an answer to see why each option is right or wrong.

1. A company runs an Amazon RDS for MySQL DB instance with one cross-Region read replica for disaster recovery. The company needs writes to resume without manual intervention if the primary Availability Zone fails. Which solution meets these requirements?

   Right. In the game that’s Renown for your Design Resilient Architectures building.

   Not that one: A is right. In the game your Design Resilient Architectures building would lose a little Renown, and this question would come back later.

   The answerA.Enable a Multi-AZ deployment for the primary, because a read replica is promoted by hand and does not provide automatic failover.Both a Multi-AZ deployment and a read replica put a second copy of the data somewhere else, and the deciding behaviour is what happens automatically. AWS describes the Multi-AZ standby's purpose as failover support: "A Multi-AZ DB instance deployment has one standby DB instance that provides failover support, but doesn't serve read traffic."Your answerB.Rely on the cross-Region read replica, which is promoted automatically when the primary Availability Zone fails and takes over the application's writes.A cross-Region read replica is not promoted automatically. Promotion is an explicit action, and the replica applies changes asynchronously, so recent writes may not have arrived.Your answerC.Create a second cross-Region read replica so that two replicas can elect a new primary between them.Read replicas are independent copies with no quorum or election between them. A second replica adds another readable copy and no failover.Your answerD.Schedule automated backups more frequently so that a restore covers the loss of the Availability Zone.Answers an availability requirement with a backup schedule. A restore is measured in minutes or hours and loses everything since the last backup.

   Show the full explanation and doc quotes

   A read replica is the other instrument: asynchronous, readable, and promoted deliberately rather than automatically, which makes it the answer to a Regional outage and not to the loss of one Availability Zone. The two are complementary, and this deployment has only the second.

   Written from docs.aws.amazon.com/AmazonRDS/latest/UserGuide/Concepts.MultiAZ.html

2. A reporting workload is overwhelming an Amazon RDS for PostgreSQL DB instance that already runs as a Multi-AZ DB instance deployment. A solutions architect must move the reporting queries off the writer without changing the application's write path, and the company wants the reporting endpoint to survive the loss of an Availability Zone. Which approach will meet these requirements?

   Right. In the game that’s Renown for your Design Resilient Architectures building.

   Not that one: D is right. In the game your Design Resilient Architectures building would lose a little Renown, and this question would come back later.

   Your answerA.Point the reporting queries at the existing Multi-AZ standby instance, which already runs in a second Availability Zone and holds the same data.Reads the standby as a warm copy that can be queried. A Multi-AZ DB instance deployment's single standby provides failover support and nothing else; it accepts no connections.Your answerB.Create a manual snapshot each morning and restore it into a second DB instance that the reporting tool queries.A daily restore is a copy of yesterday's data in a standalone instance. It neither reflects current writes nor provides failover for the reporting endpoint.Your answerC.Increase the DB instance class of the primary so that the reporting queries and the writes both fit on one instance.A larger instance still runs both workloads on the same resources. The reporting queries continue to compete with the writes, and nothing about availability changes.The answerD.Convert to a Multi-AZ DB cluster deployment and point the reporting queries at its reader endpoint.Two of these designs put a second instance in another Availability Zone, and one documented behaviour separates them. AWS distinguishes the two Multi-AZ shapes explicitly.

   Show the full explanation and doc quotes

   "A Multi-AZ DB instance deployment has one standby DB instance that provides failover support, but doesn't serve read traffic. When the deployment has two standby DB instances, it's called a Multi-AZ DB cluster deployment. A Multi-AZ DB cluster deployment has standby DB instances that provide failover support and can also serve read traffic." The company already has the first shape, whose standby cannot be queried. Converting to a Multi-AZ DB cluster gives standbys that both fail over and serve reads, and the cluster exposes a reader endpoint for exactly this traffic.

   Written from docs.aws.amazon.com/AmazonRDS/latest/UserGuide/Concepts.MultiAZ.html

3. A company runs an Amazon RDS for PostgreSQL DB instance in a single Availability Zone. The company must keep a synchronously replicated standby copy in a second Availability Zone for high availability. The company must also move its business reporting queries off the primary DB instance to a copy that accepts read-only connections. Which combination of steps should a solutions architect take to meet these requirements? (Select TWO.)

   Right. In the game that’s Renown for your Design Resilient Architectures building.

   Not that one: D,E is right. In the game your Design Resilient Architectures building would lose a little Renown, and this question would come back later.

   Your answerA.Point the business reporting queries at the standby replica of the Multi-AZ DB instance deployment.Sends reads to a copy that cannot serve them. In a Multi-AZ DB instance deployment, the standby replica does not serve read traffic.Your answerB.Create a read replica of the DB instance in a second Availability Zone, and treat it as the synchronously replicated standby copy.Misses the replication mode. RDS updates a read replica asynchronously, so it cannot be the synchronously replicated standby copy.Your answerC.Turn on read replica auto scaling for the DB instance, so that RDS adds read replicas for the reporting queries as demand grows.Relies on a feature that does not exist. Read replicas must be created manually, and RDS does not auto scale them.Your answerD.Convert the DB instance to a Multi-AZ DB instance deployment, with a standby replica in a second Availability Zone.Your answerE.Create a read replica of the DB instance, and point the business reporting queries at the read replica's endpoint.

   Written from docs.aws.amazon.com/AmazonRDS/latest/UserGuide/USER_ReadRepl.html