Why GoodData and HP Vertica have Collaborated

Why GoodData and HP Vertica have Collaborated 150 150 Kerry Butters

And, why should you care? Well, the partnership brings together two companies from opposing ends of the data analysis / management spectrum – and could light the way for future collaborations in this sector.

The Players

GoodData is a rising star in the field of cloud analytics. The venture-backed start-up recently introduced an Open Analytics Platform to provide a comprehensive set of services spanning the entire analytics lifecycle, from initial collection of records to data processing, visualisation and business intelligence.

With analysis conducted on the fly, GoodData is designed to be very easy for even non-technical business people to use.

Vertica is a high-performance columnar database which Hewlett-Packard acquired from Vertica Systems in 2011. Built to handle really large amounts of data, Vertica is better tailored to address the performance and scalability requirements of enterprise customers.

The Rationale

A key challenge faced by organisations today is the need for speed, high availability, and flexibility of data and analytics. Traditional solutions weren’t built to handle the variety and volume of data flowing through data-driven companies.

Amid increasing demand for its service, GoodData has in effect signed up as a customer of HP Vertica, which now serves as its data warehouse. GoodData claims this will lead to performance improvements, with some queries being executed orders of magnitude faster than before. Dashboards and reports should see dramatic improvements as well.

GoodData is leveraging the HP Vertica Analytics Platform to power its own cloud-based Open Analytics Platform, and enable real-time data analysis for its customers. This is done through a new data warehousing service. The package provides a single interface for scheduling, deploying and monitoring operations, and covers the gamut from load to visualisation of data.

The Way It’s Being Done

GoodData is introducing an “agile data warehousing service”. Essentially the scheme uses a columnar database alongside governance and integration processes, and includes:

·    A columnar warehouse, deployed across distributed architecture

·    Complete history of data, both in raw and transformed states, running from ingestion to visualisation

·    Integrated dual-stage storage

·    Separation of all raw data (in file storage or HDFS) from analytical areas, for changing business models

·    Data Integration Service Console for the integration of structured and unstructured data

·    Support for advanced cleansing and data manipulation

·    Fully managed clustering, provisioning, and sizing of data storage processes

·    Flexible data modelling functions for warehouse schemes, and aggregate data marts

·    Full disaster recovery and reconstruction from source data, at any time in the future

GoodData also pledges to take care of the underlying “plumbing work” (including clustering and resource allocation), while providing safeguards against data loss.

Their aim is to eliminate much of the hassle involved in managing analytical environments. The safeguards will address the reliability concerns that have traditionally held some organisations back from moving data to the cloud.

For The Customer…

With HP Vertica powering GoodData’s cloud-based Open Analytics Platform, customers are promised:

·   Fast and continuous loading from the staging and archiving system of their data storage service

·   Built-in analytical functions, and engines to increase the volume of data fed to GoodData’s analytic data marts

·   High speeds due to the pure columnar architecture, reflected in dramatic improvements in dashboard and report execution

·   Clustered, highly available, parallel architecture for infinitely scalable central warehouse storage in the long term

·   The Vertica platform, as the backbone to a complete end-to-end analytic system capable of handling all available data found within an organisation

Sounds Great. But…?

By greenlighting Vertica to be its analytics platform, GoodData is essentially admitting to the failings of its own lightweight offering.

In response to this, the company maintains that:

“We don’t sell our platform in a piecemeal fashion. Because we are a shared service hosted in the cloud, when a customer licenses GoodData, they have access to the entire platform – from back-end infrastructure (including Vertica) to our custom dashboards and reporting. We monetize through volume of data and number of users (or in the case of Powered By, number of customers), not by what part of the platform is being used.”

At a business level, companies will get an end-to-end analytics offering. And customers do want a complete solution, with all the pretty visualisations and real time analysis that can also deliver heavy-weight number crunching. Visualisation offerings from Informatica, Birst, Tableau and others offer part of the solution. But increasingly, customers want to gain a comprehensive Business Intelligence (BI) solution.

For the future, how tenable GoodData’s position is, now that it relies on HP for a significant part of its product offering remains to be seen. This is especially true, given that other vendors are doing both streaming and batch analytics from one consistent offering – within the one company