I am unable to grasp the Serverless part of DynamoDB? How can any Database be serverless?
The best I can think of is:
From a FaaS (like AWS Lamba Program, which is also serverless) one defines the schema of the DynamoDB Serverless and uses the DB/Tables created during the execution of the FaaS. Once the FaaS finishes (i.e terminates), the schema and the data of DynamoDB Serverless is also lost for ever. (So, it's something like an in-memory DB, though not exactly, which exists only for the short duration of program execution. Once the program terminates, the in-memory DB is also lost for ever. )
Is this understanding about serverless DB's like DynamoDB Serverless correct. Else please correct?
Amazon DynamoDB is offered "as a service". The infrastructure is totally operated by AWS. Applications communicate with DynamoDB via API calls and usage is charged based upon storage and provisioned throughput.
Yes, behind-the-scenes DynamoDB uses servers, but users are not exposed to this. Instead, they simply use DynamoDB. All the provisioning, networking and backup are done by AWS.