Related: DBMS
This article will use some images and diagrams that belong to: Adam Cantril - cantrill.io
Key-Value
Redis
Key must be unique.
Very simple. Scalable. Fast. In-memory caching.
/Attachments/Pasted-image-20230121045338.png)
Wide Column Store
1 Key or more keys, groupings of attributes.
Attributes can be not unique, no restriction
/Attachments/Pasted-image-20230121045357.png)
Document
Amazon DocumentDB, MongoDB, Firebase Firestore
Like an extension of Key-Value, each document interacted by its unique ID. Nested data. Each document are unique and may be linked together. Flexible index.
/Attachments/Untitled-picture-2.png)
Column
Apache Cassandra
Alike Relational Row store, column can quickly give out data for whole data in the column. Good for analytics, bad for transactions.
AWS Redshift.
/Attachments/Pasted-image-20230121045434.png)
Graph
Neo4j
Nodes and edges with name/value pairs.
Relationships are stored in the database itself, query to pull data on all people from xyz corp will run faster. Because in relation DB, relations are computed after data is taken, here not.
/Attachments/Pasted-image-20230121045429.png)