Databases & Analytics Summary in AWS
Relational database - OLTP (online transaction processing):
RDS and Aurora (SQL)
You also need to know for RDS the difference between a multi-AZ
deployment, read replicas, and multi-regions, as well as their use
cases.
1. Read Replicas
- scale the read workload of your DB
- Can create up to 15 Read Replicas
- Data is only written to the main DB
- Read Replicas are used to scale reads
2. Multi-AZ
- Failover in case of AZ outage -> high availability
- Data is only written/read to the main database
- Failover DB – in case the main database crashes, RDS will
triger a failover -> develop database in a different AZ
- The Failover DB is passive, is not accessible until there is
an issue with the main database.
3. Multi-Region
- Create a Read Replica in another Region
- Is a disaster recovery strategy
- Local performance for global reads
- Replication cost associated.
ElastiCache: In-memory database or in-memory cache.
DynamoDB, which is a serverless database & DAX (cache for
DynamoDB): Key/value database, NOSQL database.
Redshift (SQL ): Web data warehousing or OLAP (online analytical
processing).
EMR (Elastic MapReduce)service: Hadoop cluster to do big data
analysis.
Athena: query data on Amazon S3 (serverless and SQL)
QuickSight: dashboards on your data (serverless)
DocumentDB: Aurora of MongoDB (JSON – NoSQL database)
Amazon QLDB: is a financial transaction ledger (financial transaction,
immutable journal, something that is cryptographically verifiable).
Amazon Managed Blockchain: managed Hyperledger Fabric and
Ethereum blockchains
Glue: managed extract, transform and load tool, so ETL, which also
has a Data Catalog service to discover data sets into your various
databases in AWS.
DMS (Database Migration Service): Database Migration.
Neptune: graph database
Timestream: time-series database.
