MongoDB showcases serverless deployment model in MongoDB 5.0 launch

data

MongoDB has made available MongoDB 5.0, the latest release of its namesake NoSQL database platform. The new version includes enhancements such as native time series support, while MongoDB is also previewing serverless databases as part of its MongoDB Atlas cloud database service.

MongoDB 5.0 was unveiled at the firm’s MongoDB.live conference in New York. Among the most notable features added in this release is support for time series collections, which along with clustered indexing and window functions make it simpler and lower cost to operate applications such as IoT and financial analytics.

On the cloud side, MongoDB is offering a preview of serverless instances as part of its MongoDB Atlas cloud database service. This can be likened to serverless computing, which allows developers to run code without having to worry about provisioning and managing the underlying resources.  

With serverless instances on MongoDB Atlas, users get a database that dynamically adapts its resources to match their workload (up to 1TB of storage in preview), and pay only for the reads and writes the application performs and the storage resources that are actually used, the firm said. This will benefit developers by making it easier to use a database with an application, but could also lead to higher than expected charges unless a close eye is kept on the resources being consumed.

Serverless instances support the latest MongoDB capabilities (currently MongoDB 5.0) so that users never have to worry about upgrades.

Another key feature is Live Resharding, which is claimed to allow the user to change the way data is distributed across a cluster of database nodes, without incurring any database downtime. The way that data is sharded across a cluster normally has to be chosen carefully, as changing it typically involves manually exporting all the data and reloading it.

MongoDB claims that with Live Resharding, an admin can simply change the shard key on-demand as their application requires it, and MongoDB 5.0 will automatically copy and rewrite all of the current data to the new collection in the background, while continuing to accept incoming application writes.

MongoDB is also moving to offer longer-term stability with the introduction of a Versioned API. Starting from MongoDB 5.0, the Versioned API defines a set of commands and parameters that are most commonly used in applications, and the company pledges that these will not change between database releases. If changes do need to be made in the future, a new version of the API will run concurrently with the existing Versioned API on the same server, MongoDB said.

Other enhancements include MongoDB Charts integration with Atlas Data Lake. This allows users to create visualisations for data stored in Amazon AWS S3 without having to move or duplicate the data elsewhere, according to the firm.