AWS recently announced the general availability (GA) of Amazon DocumentDB Serverless, an on-demand, auto-scaling configuration for Amazon DocumentDB. However, it’s important to note that while AWS markets it as “serverless,” it aligns more with an auto-scaling model rather than a scale-to-zero model, which is a key differentiator often associated with true serverless offerings.
Amazon DocumentDB is a fully managed document database compatible with the MongoDB API (starting with Amazon DocumentDB 5.0.0 and higher). The company offers Amazon DocumentDB Serverless with the same MongoDB-compatible APIs and capabilities as Amazon DocumentDB, including read replicas, Performance Insights, I/O optimized, and integrations with other Amazon Web Services (AWS) services.
Yanny Yun, a principal developer advocate for AWS Cloud, explains:
Amazon DocumentDB Serverless introduces a new database configuration measured in a DocumentDB Capacity Unit (DCU), a combination of approximately two gibibytes (GiB) of memory, corresponding CPU, and networking. It continually tracks utilization of resources such as CPU, memory, and network coming from database operations performed by your application.
(Source: AWS News blog post)
Amazon DocumentDB Serverless is designed for a diverse array of applications with variable workloads, including multi-tenant and mixed-use (read/write) scenarios. According to AWS, this offering is particularly beneficial for enterprises that operate thousands of applications or for software-as-a-service (SaaS) vendors managing multi-tenant environments, which may involve hundreds or thousands of databases.
The company views Amazon DocumentDB Serverless as a perfect solution for sophisticated AI agents that operate with little oversight. With the unpredictable characteristics of AI workflows, demand can surge unexpectedly; hence, autoscaling can help. For instance, a travel agency using AI to organize trips might experience a sudden influx of requests during a promotional event, with countless users looking for hotel reservations, car rentals, and vacation activities all at once.
Luc van Donkersgoed, an AWS Serverless Hero, commented on LinkedIn:
AWS just released DocumentDB Serverless. Pretty cool if you have variable MongoDB workloads. Once again, they should have called it auto-scaling, though – it’s not serverless, as the minimum sizing costs $30/month.
And Tobias Schmidt, a freelance software engineer, tweeted:
Now we have a MongoDB-compatible managed database on AWS that actually auto-scales compute and memory (sadly, not to zero, though). At least no more overprovisioning for peak or sweating through surprise traffic spikes.
This approach to auto-scaling for MongoDB-compatible workloads positions Amazon DocumentDB Serverless within a competitive landscape where different cloud providers offer varying “serverless” capabilities for databases. For example, Azure Cosmos DB, with its MongoDB API, offers a serverless capacity mode that truly scales to zero when idle and bills based on consumed operations, providing a pay-per-request model without a minimum cost. In contrast, Google Cloud Platform (GCP) does not offer a native, MongoDB-compatible serverless database that scales to zero; users typically rely on third-party managed services like MongoDB Atlas running on GCP or self-managed deployments, which may offer auto-scaling but generally come with different pricing and operational models.
Lastly, pricing details and region availability are available on the Amazon DocumentDB Pricing page.
