Up until now, DynamoDB has been the only option of a truly serverless database battle-tested for production environments. Especially after launching the on-demand throughput capacity optimization, is a perfect-fit database engine for serverless projects.
The main issue with DynamoDB? Vendor lock-in.
That is one of the main reasons preventing serverless teams to adopt a serverless database. Many have falled back to SQL, by using Aurora Serverless, or to a document DB such as MongoDB. These two databases are supported by open-sourced APIs, releasing the vendor lock-in fears.
The scalability issue can be solved by Cassandra while also meeting the open standards requirement.
As an example: each node in Cassandra can serve requests directly. Without the need for a Master node, as opposed to Aurora or MongoDB, the scalability and reliability of Cassandra are increased by orders of magnitude.
Cassandra is an open-source database system and can be deployed virtually anywhere, in the cloud or on-premise. Migrating an application relying on Cassandra would be greatly faster and cheaper in comparison to one tied to the proprietary DynamoDB API.
Serverless Cassandra is still in preview mode. At Dashbird, we are testing it to really confirm it can scale to cope with high throughput of a Lambda application, for example.
We’re publishing a thorough comparison of Cassandra Serverless and DynamoDB this week. Subscribe here to receive an alert when it’s live!