MongoDB goes serverless with latest Atlas release

The NoSQL database specialist is making its managed Atlas service available on a serverless basis, and has announced a host of updates to version 5.0 of its core product.

MongoDB goes serverless with latest Atlas release
Denis Doukhan (CC0)

Popular NoSQL database provider MongoDB is jumping on the serverless computing bandwagon by offering the MongoDB Atlas database-as-a-service (DBaaS) option on a serverless basis for preview customers from today.

Serverless is an emerging cloud architecture which sees third-party providers dynamically allocate the compute and storage resources needed to execute a particular function, meaning developers no longer have to provision or manage their own servers and they only pay for when their functions execute.

“We want developers to be able to build MongoDB applications without having to think about database infrastructure or capacity management,” MongoDB CTO Mark Porter wrote in a blog post. “With serverless instances on MongoDB Atlas, now available in preview, you can automatically get the database resources you need based on your workload demand.”

For users, this means less decisions to make when launching a database. MongoDB will ask which cloud region you want to host your data in, from which Atlas will spin up an on-demand database endpoint which will dynamically flex depending on traffic levels and for which you will only pay per reads and writes.

What’s new in MongoDB 5.0

The vendor also announced the version 5.0 of the MongoDB database (you can read InfoWorld’s review of MongoDB 4.0 here), with new features including:

  • Native time series data support ­­– Time series collections, clustered indexing, and window functions have all been added to MongoDB 5.0 for a better time series data experience. In the background, MongoDB will automatically optimize schemas for high storage efficiency, low latency queries, and real time analytics. Time series collections can reside alongside other data in your MongoDB database, allowing developers to combine time series data with other enterprise data, to be queried with a single API.
  • Versioned API and live resharding – MongoDB is adding a versioned API to the MongoDB 5.0 release, meaning developers don’t need to fear that backward-breaking changes will be introduced whenever they update to the latest version, as it will now decouple your app lifecycle from the database lifecycle. The vendor is also adding live resharding, enabling changes to the shard key without downtime.
  • New privacy and security features – MongoDB now offers client-side field level encryption for added data privacy controls and the ability to reconfigure your audit log filters and rotate x509 certificates without downtime.
  • Enhanced Ops Manager controls – For MongoDB Enterprise Advanced customers, the native Ops Manager can now support the automation, monitoring, and backup/restore of MongoDB 5.0 deployments; parallelized client-side restores; a new quick start experience for deploying MongoDB in Kubernetes; and a migration assistant to push more customers towards the managed Atlas service.

Other MongoDB announcements

MongoDB is also changing its release cadence for new versions of the database, starting MongoDB 5.0. From now on the vendor will publish quarterly releases, which will roll up into a major release once a year. 

Other announcements included the general release of the redesigned MongoDB Shell, the ability to visualize Atlas Data Lake and AWS S3 data using MongoDB Charts, and new native full-text search capability within Atlas, such as function scoring and the ability to define collections of synonyms for a common search index.

Copyright © 2021 IDG Communications, Inc.