Aurora Flashcards
What are Aurora Replicas used for?
Read-scaling and failover.
What does Aurora use instead of local storage?
Aurora uses a cluster volume with a maximum of 128 TB. There are six replicas across multiple availability zones.
How is data replicated across Aurora Cluster volumes?
Synchronously.
Which instances can write to an Aurora Cluster volume?
Only the Primary instance.
How many AZs does Aurora replicate data across?
Three.
What type of storage does an Aurora Cluster volume use?
All SSD-based storage, meaning high IOPS and low latency.
How is Aurora billed?
Based on the high water mark of what you have actually used, though this will be changing.
Where does the Cluster Endpoint point for Aurora?
At the Primary instance.
Where does the Reader Endpoint point for Aurora?
At the Replica instances and the Primary instance. It load balances reads across the cluster.
What does Backtrack do for Aurora?
It allows you to rollback a database in-place to a previous point in time. No restore is needed.
How does Aurora Fast Clone work?
It creates a clone that points to the original database and then only maintains changes to the data. This makes the cloned database much smaller and much faster to provision.
What is Aurora Serverless?
A managed database service that provision Aurora Capacity Units (ACUs) from a shared pooled. ACUs can be scaled up or down based on need and limited by rules for the cluster.
When should you use Aurora Serverless?
- Infrequently used applications
- New applications that you are unsure about
- Variable workloads
- Unpredictable workloads
- Development and test databases
- Multi-tenant applications
When would you use Aurora Global Database?
- Cross-region disaster recovery and business continuity.
- Global read scaling - ~1 second latency for replication across regions.
What Aurora mode allows multiple read/write instances?
Aurora Multi-Master.
