AWS Fundamentals: Amazon Aurora

Question 1

What is Amazon Aurora?

Answer 1

Aurora is AWS’s proprietary relational database service compatible with MySQL and PostgreSQL. It is designed for cloud-native applications with high availability, scalability, and performance improvements over standard MySQL and PostgreSQL.

Question 2

What performance improvements does Aurora provide?

Answer 2

Aurora provides 5x performance improvement over MySQL and 3x over PostgreSQL on RDS, thanks to optimizations specific to the cloud.

Question 3

What is the storage system in Aurora?

Answer 3

Aurora automatically scales storage from 10GB to 128TB without manual intervention, and it is stored across 3 availability zones with self-healing and replication features.

Question 4

How does Aurora ensure high availability?

Answer 4

Aurora replicates data across 6 copies in 3 AZs, requiring only 4 copies for writes and 3 for reads, ensuring high availability even if one AZ fails.

Question 5

How fast is Aurora’s failover process?

Answer 5

Aurora’s failover takes less than 30 seconds, ensuring minimal downtime.

Question 6

How many read replicas can Aurora support?

Answer 6

Aurora can have up to 15 read replicas, which are used to scale read workloads.

Question 7

What are Aurora’s two main endpoints?

Answer 7

Writer Endpoint: Always points to the master instance for writes.

Reader Endpoint: Automatically load-balances connections to read replicas.

Question 8

What makes Aurora’s storage different from traditional databases?

Answer 8

Aurora uses a shared storage volume across multiple AZs, which is auto-expanding, self-healing, and replicated for reliability.

Question 9

What is the Aurora Backtrack feature?

Answer 9

Aurora’s Backtrack allows you to restore data to any specific point in time, without relying on backups, providing a flexible recovery mechanism.

Question 10

What is the cost difference between Aurora and RDS?

Answer 10

Aurora is about 20% more expensive than RDS but provides greater efficiency at scale, making it cost-effective in large applications.

Question 11

What makes Aurora auto-scaling for read replicas unique?

Answer 11

Aurora can auto-scale the number of read replicas based on demand, ensuring that there are always enough read replicas to handle the load.

Question 12

How does Aurora handle replication lag?

Answer 12

Aurora typically experiences sub-10 ms replication lag for read replicas, making it much faster than standard MySQL replication.

Question 13

How does Aurora handle security and compliance?

Answer 13

Aurora provides built-in security, including encryption, and complies with industry standards for security and data isolation.

Question 14

What is a key advantage of Aurora’s architecture?

Answer 14

Aurora’s architecture is highly resilient and scalable, with automatic failover, automated backups, and self-healing capabilities that reduce the risk of downtime.

Question 15

What is Replica Auto Scaling in Amazon Aurora?

Answer 15

Replica Auto Scaling automatically adds Aurora replicas when read traffic increases, helping to distribute the load across multiple replicas and reduce CPU usage on the primary instance. The Reader Endpoint will be updated to include the new replicas.

Question 16

What is the purpose of Custom Endpoints in Aurora?

Answer 16

Custom Endpoints allow you to define a subset of Aurora instances (based on instance types) for specific workloads. This is useful for directing analytical queries to more powerful replicas, and it may replace the use of the Reader Endpoint.

Question 17

What is Aurora Serverless and when is it useful?

Answer 17

Aurora Serverless automatically scales Aurora based on actual usage. It’s ideal for intermittent or unpredictable workloads where capacity planning is difficult. You pay by the second, making it more cost-effective.

Question 18

What is Aurora Global Database?

Answer 18

Aurora Global Database allows for cross-region replication with one primary region for reads and writes and up to five secondary read-only regions. It supports disaster recovery with an RTO of less than one minute and replication lag of less than one second.

Question 19

How does failover work in Aurora Global Database?

Answer 19

In case of a failure in the primary region, Aurora Global Database allows for quick failover to a secondary region. The secondary region can be promoted to a read-write cluster in less than one minute.

Question 20

What is Aurora Machine Learning and how does it work?

Answer 20

Aurora Machine Learning integrates with AWS ML services (like SageMaker and Comprehend) to perform ML-based predictions directly from SQL queries. Use cases include fraud detection, ads targeting, sentiment analysis, and product recommendations.

Question 21

How does Aurora work with SageMaker?

Answer 21

Aurora can send data (e.g., user profile, shopping history) to SageMaker via an SQL query. SageMaker then processes the data and returns predictions (e.g., product recommendations) to Aurora, which passes the results back to the application.

Question 22

What is the replication lag for Aurora Global Database?

Answer 22

Aurora Global Database provides less than one second replication lag across regions.

Question 23

How does Aurora handle backups and recovery?

Answer 23

Aurora has automated backups and supports point-in-time recovery. It performs continuous backups to Amazon S3, enabling fast recovery in case of failure.