Let’s summarise your learnings from this session. In the next video, our SME will provide a brief summary of all the concepts you learnt in this session.
Some of the points from the video above are summarised below.
- Redshift is based on the massively parallel processing architecture.
- In Redshift, data is stored in the columnar format. This offers superior performance for analytical queries.
- Redshift does not enforce constraints such as unique key, primary key and referential integrity. However, the optimiser uses it to create a query plan.
- The sort keys determine the order in which the data on the slices will be sorted.
- Unlike traditional databases, Redshift does not use indexes; instead, it uses zone maps.
- The individual columns in Redshift tables can be compressed using compression encodings. There are 13 different encodings available in Redshift.
- Since there is some amount of overhead sorting and distribution while writing rows into tables, Redshift is not suitable for OLTP workloads.