Couchbase Capella gets columnar support on AWS

NoSQL document-oriented database provider Couchbase is adding columnar support to its managed database-as-a-service (DBaaS), Capella, on AWS.

Here columnar support refers to the columnar database management system, which is typically NoSQL in nature and is implemented for complex analytical tasks and queries.

In contrast to relational databases, columnar databases or database management systems store data in columns instead of rows. This is done in order to reduce storage costs and reduce the time taken to return a response to a query, thereby improving performance.

Apache Parquet, MariaDB, ClickHouse, and Apache Cassandra are examples of columnar databases.

Columnar support, which has been made generally available, has been added to Capella in order to address enterprises’ need to analyze dormant semi-structured JSON data, according to the company.

Columnar support to help analyze dormant JSON data

The semi-structured JSON data remains dormant at most enterprises because it can be difficult to use for traditional analytic systems that expect data to conform to more rigid structures, Couchbase explained, adding that without formal structures, business intelligence (BI) teams may have to spend too much time on data hygiene, resulting in abandonment of the data for analytics.

This means that Capella now offers key-value and columnar storage options for operational and analytic workloads on a single platform, making it easier for developers to use JSON data for analytics.

The dual support for database storage, according to Amalgam Insights’ chief analyst Hyoun Park, will help Capella meet the present demand of enterprises as well as compete with rivals.

“Couchbase has long been a database solution of choice for processing semi-structured data, but digital companies increasingly need to support a combination of analytics, predictive AI, search, generative AI, and real-time data processing, which are all often strung out across multiple data platforms. Capella’s columnar support addresses that demand,” Park explained.

In addition, the chief analyst said that the addition of columnar support to Capella will help Couchbase bridge the “analytic” gap that previously separated it from MongoDB Atlas and Google BigQuery.

“Couchbase has typically been seen as easier to use because of its flexibility of data and horizontal scaling, but that flexibility came at the price of only supporting operational data access rather than more complex analytic processing use cases,” Park explained, adding that this opens up avenues for Couchbase’s existing customers to simplify their database portfolio. 

The addition of the new feature may also help Couchbase expand and target larger enterprises, according to Alexandar Wurm, senior analyst at Nucleus Research.

“Columnar support is important for operations that involve aggregating, summarizing, or searching across large datasets. This becomes especially important at enterprise scale,” Wurm explained.

Capella iQ to help with analytics

The new columnar support feature parses and transforms JSON data into an analysis-ready columnar format to help with analytics, according to the company.

It uses Capella iQ, a proprietary generative AI-based coding assistant built into the DBaaS’ Workbench, to help any enterprise developer run analytics-related tasks.

The generative AI-based assistant, which was released last year, typically can help developers write SQL++ queries, build indexes, define text searches, write application-specific code, and create functional test cases.

Further, Capella’s columnar storage supports real-time, multi-source ingestion of data not only from Couchbase but also using common systems like Confluent Cloud — built by the original creators of Apache Kafka — to draw data from other third-party JSON or SQL systems.

The company has confirmed that columnar support for Capella will be released on other public cloud service providers, such as Microsoft Azure and Google Cloud.

However, Park pointed out that Couchbase may have chosen AWS to start with as it is the largest starting point from a market perspective and each cloud-specific launch is effectively a new product that has to be supported in the context of each hyperscaler’s technology stack, geographic footprint, and service roadmaps.

Capella gets a free tier and other updates

Couchbase is also planning to introduce a new free tier, without any time limits, for developers to try out Capella.

Developers will be able to try out Capella along with its features, such as Capella iQ and Capella Workbench, the company said, adding that the free tier will be available in September.

Further, developers will be able to use integrated development environments, such as Visual Studio Code along with community-supported extensions, such as cbshell 1.0, which supports vector search, and plugins such as Ionic Capacitor for Couchbase Lite, which simplifies cross-platform mobile development.

Capella, currently, has three plans for its subscribers — Basic, Developer Pro, and Enterprise — priced at $0.15 per hour per node, $0.35 per hour per node, and $0.56 per hour per node respectively.

In other updates, Couchbase said that it has started offering vector support to its Couchbase Lite database which is targeted for mobile and internet-of-things (IoT) applications.

The feature, dubbed Couchbase Mobile with Vector Search, allows developers working on developing applications for mobile devices to leverage vector search at the edge for semantic search and RAG tasks without being connected to the internet, the company said. This could help enterprises using AI in their mobile applications reduce downtime due to internet outages, it explained.