Capella adds Columnar Analytic Services to GCP

Couchbase’s Capella Columnar, the analytic services of our developer data platform, is coming soon in Google Cloud, expanding its support for popular clouds. The Analytic Services include real-time ingestion and conversion of complex JSON data (aka ZeroETL) from Capella’s Transactional Services and Kafka-based ingestion from third-party databases and applications.

It includes massively parallel columnar processing separated from its storage features, query without creating index, Capella iQ natural language-powered query and visualization, BI tool integrations, and real-time writeback to Capella’s platform services and Couchbase Server.

These analytic services address long-standing issues faced by developers and data scientists in incorporating semi-structured JSON data in their analytics, data science and AI-base applications.

Capella Columnar Analytic Services on Google Cloud

Analytic Services use optimized GCP instances

The GCP deployment of Analytic Services supports Compute-Optimized (C4A) instances featuring Axion Arm-based processors and Titanium SSDs. This environment offers exceptional performance for both the compute and storage needs of Capella.

Analytic Services and GCP are ideal for AI-powered applications

“Developing AI-based applications and experiences is going to become the next great enterprise brand distinction. The combination of Couchbase’s Capella Columnar running on Google Cloud’s Axion-based instances creates instrumental value for the future’s critical, AI-powered (or agentic) applications,” said Matt McDonough, VP of Products and Partnerships.

“Our customers regularly choose Capella for its performance, versatility and scale as the database foundation for their analytical applications. And they will undoubtedly appreciate the exponential boost in performance and value of Arm-based Axion infrastructure for these workloads, especially when they need to offset the costs of incorporating AI within them. The analysis will show that this is where AI’s budget will be revealed.” – Salil Suri, Director of Product Management, Google Compute Engine

About Google’s Arm-based Axion instances

Axion is built on the standard Armv9 architecture and instruction set. Axion sets a new bar in cost effectiveness with both higher raw performance and lower energy consumption over historic architectures. Match that with Capella’s Analytic Services high-speed columnar processing and they create an unprecedented value for customers–faster real-time analytics running on faster C4A instances, while consuming less electricity and operating with lower costs is an enterprise dream solution. Capella’s Columnar Analytic Services on Google Cloud, brings a new era of performance to JSON data analytics. Built on Google’s C4A instances with Arm-based processors and Titanium SSDs, Columnar addresses the historical challenges of incorporating JSON data into analytics, machine learning, and AI, enabling developers to build cutting-edge AI-powered applications without breaking the bank.

Analytic Services expands Capella’s capabilities on GCP

While Capella’s traditional Transactional and Mobile services have been available in Google Cloud for years, adding Analytic Services allows customers to enjoy and add real-time analytic processing within the same GCP environment.

Common Capella interfaces and functionality on GCP

Customers will also recognize familiar capabilities that are common to the Capella platform for both development and operations, including:

    • Operating experience through a common user interface, security, operations, and billing
    • SQL++ is our common query language across the platform and natural language processing is available from Capella iQ with extensions for building analytic visualizations using iQ Insights
    • In-memory data processing and communication to facilitate both transactional and analytic activities
    • SDK-based development tools and data API

Congruent capabilities across Capella platform services in GCP

While Capella’s platform services share common development and DevOps characteristics, they also provide separate, but congruent, features that differentiate their data access, processing, transit and storage capabilities for their roles in transactional, analytical, mobile and AI operations. For example:

    • Capella’s transactions services match those of Couchbase Enterprise Server for data access via key/value, document, search, vector search, time-series and graph relationship techniques, however Capella’s Analytic Services offer data access via interfaces popular with business intelligence tools like Tableau and PowerBI
    • Capella’s platform services store its data in different underlying persistent stores that are tailored to the performance and scaling requirements of their respective roles as transactional or analytic stores
      • Magma: a high-performance, distributed key-value engine, persists data for our transactional services
      • Couchbase Lite: the device-side persistent store for our Mobile Services
      • Analytic services are stored in columnar structures within GCPs object storage services
    • Data movement systems including replication, synchronization, ingestion, and writeback are also processes that are unique to transactional, mobile or operational services

Capella’s Columnar Analytic Services offer breakthrough features on GCP

Zero ETL for JSON and semi-structured data

Couchbase knows and accommodates JSON application data that evolves over time. This simplifies data design and implementation without breaking the application. The same holds true for our Columnar Analytic Services, where users don’t need to define a schema in advance. It automatically handles new fields, mixed data types, nulls, and missing values—allowing customers to focus on the query without the extra hassle of evolving and maintaining detailed schema.

Enterprise data requires multi-source Ingestion

Analytic Services support heterogeneous data ingestion from NoSQL, RDBMS, or flat file sources. Real-time data is automatically ingested from Capella’s transactional services, and uses Kafka as its real-time ingestion pipe from external data sources like MongoDB. It can also process semi-structured files including CSV, TSV, Avro, Iceberg and Parquet formats.

Operational write-back

Operational write-back drives further personalization, promotion and action from calculated metrics delivered as they are calculated to Capella’s Transaction Services. This will be especially useful for including real-time calculated metrics as variables for GenAI prompts.

Works in GCP’s open ecosystem

Build Operational/Analytic applications within the Google Cloud’s Open Ecosystem including development tools, Gemini-powered AI, CNCF-centric deployment and operational tools.

Powerful and economical

Our Analytic Services support separation of compute and data storage, for unlimited data set sizes, and its serverless MPP engine scales independently–all of which is optimized on GCP Axion instances. The result is extremely scalable, yet affordable pay-as-you-go operations.

 

Capella Columnar Analytic Services shine on Google Cloud

Demand within the Couchbase community has been extremely high for Google Cloud support. Now that it has arrived, customers will enjoy the following benefits:

    • Better scale and performance
    • Deployment flexibility
    • Simplified operations on GCP
    • Access to Google AI services
    • Real-time embedded AI applications
    • Unified enterprise analytics across all operational and analytic systems
    • Secure operations through Google Cloud and Couchbase

For more information on Capella Analytic Services, check out the Google Cloud Marketplace or go to the Capella Columnar information page.

Author

Posted by Jeff Morris, VP Product Marketing

Jeff Morris is VP of Product and Solutions Marketing at Couchbase. He's spent over three decades marketing software development tools, databases, analytic tools, cloud services, and other open source products. He'd be the first to tell you that anyone looking for a fast, flexible, familiar, and affordable cloud-to-edge database-as-a-service can stop looking after they check out Couchbase.

All Posts

Leave a reply