Amazon DynamoDB: keys, queries, capacity and backups

Since Baobab Market's shop has been running on several instances, carts disappear when a customer switches instance. Moussa would like to install a database on a server; Awa wants a serverless database that handles the sales and costs almost nothing while the shop sleeps. This fifth lab of the preparation path for the AWS Certified Cloud Practitioner certification starts with what the exam asks most often about databases: choosing between relational and NoSQL, between RDS, Aurora, DynamoDB, ElastiCache, Redshift, DocumentDB and Neptune, and knowing what remains your responsibility depending on the service. Then you design a DynamoDB table from its questions, on a diagram: partition key, sort key, partitions copied across three zones. You create the table after a first denial triggered on purpose, write orders with different attributes into it, then compare a query and a scan: the console shows the items read, returned, and the read units consumed. You query and update the table in PartiQL, choose between on-demand and provisioned capacity, then protect the data: automatic expiry of carts, restore to the second, a backup before the sales, encryption always on. Everything happens in the real AWS console. The guide is illustrated with screenshots of the console and a diagram, explains global tables and DAX, and ends with a ten-question practice exam.