chapter 9 Flashcards
What are the differences between relational and nonrelational databases?
Relational databases are designed for structured data that contains a defined number of attributes per record and performing complex queries.
Nonrelational databases are designed for data that doesn’t follow a predictable structure and is based on a primary key search.
What does the vertical and horizontal scaling options do for RDS?
Vertically, you can upgrade to a larger instance class to give it more processing power, memory, or disk or network throughput.
For horizontal scaling of reads, your only option is to use read replicas.
How do you describe the components of RDS?
An RDS deployment consists of at least one instance.
Must select an instance class that defines the vCPUs and memory for the instance.
Must also select a database engine.
For storage, you must select general-purpose or provisioned IOPS SSD.
You can also add read replicas to scale horizontally to improve read performance.
In a multi-AZ deployment, you can add additional secondary instances that the primary synchronously replicates data to.
What are the the backup and recovery options for RDS?
You can schedule automatic snapshots for your RDS instance to occur daily during a 30-minute backup window of your choice.
Enabling automatic backups also enables point-in-time recovery, allowing the restoration of a failed database up to 5 minutes prior to failure.
Restoring from a snapshot entails creating a new instance from the snapshot.
How does DynamoDB store data?
DynamoDB stores data as items in tables.
Each item must have primary key whose values are unique within the table.
The primary key’s name and data type must be defined when the table is created. When you create an item, you can also add other attributes in addition to the primary key.
The number of partitions allocated to a table depends on the number of WCU and RCU you configure.
What is the primary scenario for using Redshift?
Redshift is a data-warehousing service for storing and analyzing structured data from multiple sources, including relational databases and S3. Redshift can store much more data than RDS, up to 2 PB!
