Cloudera introduces PaaS built with SDX

Cloudera Altus

Cloudera recently announced Cloudera Altus with SDX, the first machine learning and analytics Platform-as-a-Service (PaaS), built with a shared data catalog providing the business context of that data. Altus supports a variety of high-value business use cases that require applying multiple data analysis capabilities and approaches together. SDX makes it possible for those analytic functions to work together to combine data from different sources into a single coherent and actionable picture. Example use cases include answering complex questions about customer “next-best-offer”, IoT predictive maintenance, and advanced threat detection.

SDX enables Altus cloud services – including Data Engineering, Analytic Database (beta) and soon Data Science – to securely access data through a reliable shared data experience. There is one trusted source of metadata for all machine learning and analytics services and users. Altus brings the simplicity and scale of the cloud to big data analytics, enabling people to confidently utilize multiple analytics services to unlock the value in their business data. Altus delivers IT control through simplified workload management, governance, and security, while catering to the end user by offering curated self-service access to data and their preferred tools.

Industry analysts have identified big data analytics cloud sprawl as a growing problem in enterprises. The disparate cloud services spun up as “shadow IT” by different teams present IT and organizational challenges because the discrete models and fragmented approaches are usually too narrow and not scalable to manage within the company. It also leads to increased cost, effort, and compliance challenges associated with ungoverned data replication and access.

According to IDC MarketScape, Asia/Pacific Big Data and Analytics Platform 2017 Vendor Analysis, “In 2017 and beyond, IT buyers, which include the various LOBs considering investing in big data and analytics and cognitive computing, would have to consider more than just a single use case within their respective business units. BDA (Big Data Analytics) has been well established on the ROIs and relative ease at which each individual business unit is able to adopt a BDA solution and rapidly apply it within their environment. The common challenge faced is when attempting to scale or replicate success achieved to more LOBs or function groups.”

Well-known data analytics services fall short of being able to solve these business use cases because they support only single-purpose workloads that are not designed to work with other cloud data services. Often data must be manually moved from service to service, stored in a new format, and have usage policies re-defined at great effort and risk. Attempting to integrate these analytics services requires expensive custom development, and may result in poor performance, inconsistent user experience, and security and governance problems.

“Cloudera Altus with SDX enables businesses to build and manage multi-function analytics use cases in the cloud, integrating data engineering, IoT, customer and operations analytics, with machine learning,” said Vikram Makhija, general manager, Cloud Business Unit, at Cloudera. “Cloudera offers a proven solution for businesses to capitalize on the value of their data, avoiding the analytics cloud sprawl problem through the simplicity and scale of Cloudera’s modern cloud platform for machine learning and analytics.”

Cloudera SDX, introduced in September 2017, makes multi-function data use cases easier to develop, less expensive to deploy, and more consistently secure. SDX is a modular software framework that applies a centralized, consistent framework for schema, security, governance, data ingest and more, making it possible for dozens of different customer applications to run against shared or overlapping sets of data. Now SDX, currently available as a self-managed reference architecture for Cloudera Enterprise, will also be available in Cloudera Altus, making it even easier for organizations to build high value multi-function data use cases.

Altus: Enhanced and Expanded

Altus runs on Amazon Web Services (AWS) infrastructure, with support for Microsoft Azure infrastructure in beta. The Altus cloud service offerings include:

Altus Data Engineering is a jobs-focused platform to facilitate ETL and data preparation for analytics and data science in cloud. It simplifies resource allocation, job creation, and troubleshooting for users. It is part of a more tightly integrated horizontal PaaS to support diverse analytics and data science use cases. The Altus SDK for Java allows users a means to programmatically leverage a platform-as-a-service for data engineering workloads.

Altus Analytic DB (beta) is the first data warehouse cloud service that brings the warehouse to the data through a unique cloud-scale architecture that eliminates complex and costly data movement. It delivers instant self-service BI and SQL analytics to anyone, easily, reliably, and securely. Furthermore, by leveraging SDX, the same data and catalog is accessible for analysts, data scientists, data engineers, and others using the tools they prefer – SQL, Python, R – all without any data movement.

Altus Data Science (beta soon) provides data science teams with on-demand Python and R services for advanced analytics and machine learning. With a serverless user experience, data science teams spend more time delivering value, and less time on DevOps. It works on a common framework of services for security, governance, data ingest, and data cataloging. This makes it easy to integrate data science with analytic database and data engineering functions, for a complete analytic solution.

“Our customers get an easier, more unified, and enterprise-grade approach to deploying and managing cloud-based analytics through our partnership with Cloudera,” said Nick Halsey, chief executive officer of Zoomdata. “Together, we will continue helping customers with a shared big data analytics experience across all deployment types, including multiple public, private, and hybrid cloud, as well as bare metal configurations.”