Intro

• SQL databases (RDS & Aurora) are centralized and can scale reads by adding read replicas but they cannot scale writes (no sharding). The only way to scale writes is to vertically scale the DB.
• DynamoDB is a Serverless NoSQL DB with replication across multiple AZ. It is a distributed NoSQL DB that can scale both reads and writes horizontally to support massive workloads.
• Database is already created and managed by AWS. We only need to create Tables.
• Table classes
  • Standard
  • Infrequent Access (IA)
• Popular option to use as a serverless distributed cache. Latency will be higher than ElastiCache as DynamoDB is not an in-memory store. It stores data on disk.
• Copy a table across account or region (3 options):
  • Using AWS Data Pipeline
  • Backup and restore into a new table (takes some time but easy)
  • Scan and BatchWriteItem (not recommended)
• Global Tables - multi-region, multi-active, replicated tables (need to enable DynamoDB streams)
• DynamoDB Local - deploy DynamoDB locally for development or testing

Tables

• Primary key must be decided at creation time
• Infinite rows (items)
• Attributes (columns) can be added overtime
• Maximum item size: 400 KB
• Supported data types:
  • Scalar - String, Number, Binary, Boolean, Null
  • Document - List, Map
  • Set - String Set, Number Set, Binary Set

Primary Keys

• Partition Key (Hash)

  • Partition key is hashed to find the location of data
  • Partition key must be unique for each item

  

• Partition Key (Hash) + Sort Key (Range)

  • Partition key is hashed to find the partition and sort key is used to search for the data in that partition
  • Combination of partition key and sort key must be unique