Understanding the Differences Between Serverless and Non-Serverless NoSQL Databases: Pros, Cons, and Use Cases

?? Hello there! Welcome to The Serverless Spotlight!

In this week's edition, I'll explain the subtle differences between a serverless NoSQL database and its non-serverless counterpart. I'll also lay out the pros and cons for each and recommend when to use which.

A constant concern for backend developers today is choosing between serverless and traditional services.

Serverless functions (Lambda) vs traditional server instances (EC2).

Serverless storage (S3) vs block storage (EBS).

Serverless container computing (Fargate) vs traditional container orchestration (ECS).

Serverless messaging queues (SQS) vs traditional message brokers (RabbitMQ).

Serverless databases (DynamoDB) vs traditional NoSQL databases (MongoDB).

The last one is the concern of this article.

What Is A Serverless Database?

A serverless database NoSQL is a cloud-native database system where the infrastructure and operational tasks of backups, maintenance, and server patching (updates), are fully managed by the cloud provider.

Another principle element of a serverless database is that scaling is usually fully or partially managed by the database system.

In the case of DynamoDB, scaling high does not take more effort from the user than proper key design. Complex tasks like data replication, sharding and horizontal scaling of your database is handled automatically for you.

With a serverless database, you can automatically scale based on demand and be charged only for the capacity you use and nothing more (contrary to non-serverless databases).

This makes serverless databases highly cost-effective for various workloads (when used correctly).

Key features of a serverless database

Some key features of a serverless database include:

• no server management
• fully managed operations (like updates, backups, etc)
• auto-scaling capabilities
• pay-per-use (pricing is based on the amount of data read/written or stored)

What is a non-serverless database?

A non-serverless database is a database where the developer is responsible for provisioning, managing, and maintaining the underlying database infrastructure.

You need to update the system when necessary, manually backup data, and scale it up or down yourself.

While you get more fine-grained control of the database system, this brings in more complexities - complexities that are oftentimes not necessary for most to most users to be responsible for.

The bottom line here is to understand your needs well and carefully consider when you need to be able to configure the underlying servers and scaling capabilities and when you don't need to.

Pros and Cons of Serverless Databases

Additional benefits of serverless databases include being able to get to market quicker. Being able to focus more on the data lets you build faster.

The fact that they are more cost-effective lends an edge to your business goals.

Some cons include:

• Limited customization: As mentioned earlier, non-serverless databases give you access to the database's underlying server (as with traditional SQL databases). With serverless databases, this is usually not possible.
• unpredictable costs: while you only pay for what you use, sometimes your application's traffic patterns may become unpredictable, leading to unforeseen spikes in costs.
• Vendor lock-in: serverless databases often lock you into the cloud provider's ecosystem, and limit your integration possibilities and flexibility

Pros and Cons of non-serverless Databases

Some pros of non-serverless databases include:

• Full control over infrastructure.
• More predictable performance (since there is no sharing of resources).
• Custom optimizations.

The cons:

• operational complexity (you need to manage infrastructure, scalability, patching, updating servers, and maintenance).
• Pay for what you don't use (with non-serverless you pay for idle resources).
• Harder to scale (requires you to manually scale your database)
• Higher maintenance overhead and costs.

Use Cases for either database types

What are some great use cases for serverless databases?

I've worked with many databases in production, both serverless and non-serverless types. Here are my recommendations based on both good practices and experience:

• event-driven architectures: serverless databases work really great with event-driven architectures and microservices.
• Startup or simple applications: for early-stage businesses or simpler apps that will have unpredictable traffic patterns, serverless databases work really well. The ease of management, flexibility, and low costs are quite beneficial for these types of workloads.
• When you do not have the expertise to manage servers and infrastructures and need to get started with a lean team.
• Small applications that need quick setups
• Applications that need the database to scale massively (beyond most traditional databases)

What are some good use cases for non-serverless databases?

• strict compliance requirements: Some applications (like healthcare or financial ones) require full control over the infrastructure.
• large businesses: some large companies need steady, predictable workloads and usually prefer to be able to control and manage the underlying infrastructure of the databases. Sometimes they also want to avoid vendor lock-ins.

Conclusion

Both serverless and non-serverless databases offer their benefits.

Like most things, the decision to use one or the other often lies in the use case of the application.

Serverless databases are advantageous in environments where scalability and cost-efficiency are critical, especially for variable workloads or small teams.

On the other hand, non-serverless databases provide more control, consistency, and predictability, making them ideal for large-scale enterprise applications or those with strict compliance needs.

?? My name is Uriel Bitton and I hope you learned something in this edition of The Serverless Spotlight

?? You can share the article with your network to help others learn as well.

?? If you want to learn how to save money in the cloud you can subscribe to my brand new newsletter The Cloud Economist.

?? I hope to see you in next week's edition!

References

1. Traditional databases vs serverless databases. Prisma.io https://www.prisma.io/dataguide/serverless/traditional-vs-serverless-databases
2. What is a cloud database? An in-depth cloud DBMS guide. Sean Michael Kerner. https://www.techtarget.com/searchdatamanagement/definition/serverless-database