Item: Amazon DynamoDB
Rating: 10
Author: stefan steinike

Use Cases and Deployment Scope

DynamoDB is our primary choice for cloud databases when relational structures are not required. The Key Value pair model is easy to interact with and strictly typed. We have many use cases that rely on DynamoDB. It holds audit records, global metadata, tracks systems statuses, and also acts as an action trigger when combined with Streams. Currently, Streams is what makes it so powerful in our current configuration. Having Streams throw off a time-ordered sequence of events for use with other services within AWS is extremely versatile.

Pros and Cons

Extremely fast
Lightweight Key-Value pairs
Serverless
Streams

On-Demand vs Provisioned capacity costs
No relational support
Limited query options

Return on Investment

Saved 30% for On Demand low-use table read costs
Streams enables triggers with other native services like Lambda
No DB administration required
No security concerns

Scalability

DynamoDB is natively designed to scale, and from a physical perspective, being on SSD's by default makes it highly performant. You can configure DynamoDB to scale based on certain criteria/metrics or stick with basic On Demand usage. There are no table size limits and no need to reserve or predict future size requirements. We have yet to see any issues where DynamoDB was unable to meet the demands placed on it.

Performance

We have yet to encounter any performance issues from DynamoDB regardless of the level it has scaled to, or the region it is deployed into. As mentioned previously, DynamoDB data is physically on SSD's across multiple availability zones so performance has always been very consistent and very fast. Dynamo also offers DAX if the default performance (milliseconds) is not meeting business needs.

Alternatives Considered

Amazon Relational Database Service (RDS)

Comparing RDS and Dynamo is not fully Apples to Apples comparison. RDS is a more flexible cloud-native solution that supports a wide range of engines that are relational. It is great for running older DB types like Oracle in the Cloud. Because it supports multiple engines, it is a popular choice for doing hybrid design between Cloud and On-Prem.
DynamoDB is a Key Value database option, which shares some of the performance and security that RDS does with a more specific focus/use case.

Key Insights

Do you think Amazon DynamoDB delivers good value for the price?

Yes

Are you happy with Amazon DynamoDB's feature set?

Yes

Did Amazon DynamoDB live up to sales and marketing promises?

Yes

Did implementation of Amazon DynamoDB go as expected?

Yes

Would you buy Amazon DynamoDB again?

Yes

Other Software Used

AWS Lambda, AWS Organizations, Amazon S3 (Simple Storage Service)

Likelihood to Recommend

DynamoDB is highly scalable and fast. It automatically puts your tables in multiple (3+) availability zones on physical SSD's. This makes it good for our critical and latency-sensitive DB deployments. It can be configured for specific demands/workloads, if needed. It offers Exports to S3 and Streams which are very powerful additions.
It is obviously not a good choice when seeking any type of relational or multi-table work, where joining might be required. It is also not a good choice for doing hybrid work between cloud and on-prem.

DynamoDB Review

Overall Satisfaction with Amazon DynamoDB