CloudGuild · Blog · Cheat sheets · Lessons · Certifications
Amazon Redshift — Fast, fully-managed data warehouse
Amazon Redshift is a fully-managed, petabyte-scale data warehouse service designed for analytics.
What it is
- A fully-managed data warehouse service for big data analytics.
- Uses columnar storage and parallel processing to optimize query performance.
When I reach for it
- When needing to analyze large datasets quickly and cost-effectively.
- For business intelligence (BI) applications and complex queries on structured data.
- When integrating with other AWS services, like S3 for data loading and AWS Glue for ETL.
Key architectural decisions
- Choose between standard and dense storage types based on data patterns and query needs.
- Decide on the number of nodes and node types based on expected workload and performance requirements.
- Implement data distribution styles (KEY, EVEN, ALL) to optimize query performance.
Gotchas & exam traps
- Understand that Redshift is not a transactional database (OLTP); it’s optimized for OLAP.
- Be aware of the need to manage vacuuming and analyze operations to maintain performance.
- Know that Redshift Spectrum allows querying data in S3, but there are additional costs involved.
The architect view
- Redshift is ideal for organizations looking to scale analytics without the overhead of managing infrastructure.
- Focus on optimizing schema design and distribution styles to enhance performance.
- Regularly monitor and adjust cluster configurations based on usage patterns.