Many of the microservices in our VOD and Replay platform use DynamoDB as their database.
Performance is very good if the data is architected for it, scalability is reasonably fast, and the serverless aspect offloads a lot of the administration and hosting work. Whether it’s performance, resilience or time-to-market, DynamoDB helps us achieve our business goals.
That said, when we spend several hundred thousand dollars on DynamoDB every year, any optimization is good for us!
With DynamoDB, committing to a certain capacity for a year can help reduce costs – up to 50% savings on that capacity. But how do we know how much to reserve when traffic on our platform varies throughout the day?