Cloud-based database software adds new IT options for users


Oracle Database Cloud Service aims to handle apps of all sizes

Oracle continues to take the cloud fight to AWS and other more-established rivals as it expands support for running web-scale and enterprise applications on its databases and IaaS platform.

Oracle continued its push into the cloud this week -- with enhancements aimed primarily at existing customers looking to run at least some of their Oracle databases and applications on cloud systems.

At the Oracle CloudWorld conference in New York, the company made its Oracle Database Cloud Service available in a bare-metal implementation running on dedicated hardware and storage. It also introduced new virtual-machine (VM) compute, load balancing and storage features that, along with the expanded database service, aim to increase the ability of Oracle's infrastructure as a service (IaaS) platform to support large-scale enterprise and web applications.

Oracle said its databases running on bare-metal, non-virtualized infrastructure can better meet the performance needs of I/O-intensive e-commerce, transaction processing and analytics jobs. That and other elements target the Oracle cloud database technology squarely at the vendor's current installed base, according to IDC analyst Larry Carvalho.

"For non-Oracle shops, it will be a tough sell in the current state," Carvalho said. But he thinks that in the long term, if Oracle can add a variety of easily deployable cloud services, it could open up its offerings to more organizations that see potential benefits in Oracle’s integration of platform services and analytics functionality.

While the base functionality in Oracle Database Cloud Service isn't an exceptionally innovative addition in an already competitive field, Oracle may stand out for connecting its software as a service applications to its IaaS platform to enable customers to build applications quickly, Carvalho said.

He added that Oracle has a good roadmap but is still relatively new in the cloud database field compared with its competition. Although the company has a chance to differentiate itself, Carvalho said time is short outside of the Oracle installed base because of the market lead that Amazon Web Services and Microsoft have gained with their Amazon Relational Database Service and Azure SQL Database offerings, respectively.

One problem with Oracle Database Cloud Service is the lack of an easily digestible overview of what exactly it offers to users, said Bob Sheldon, a technical consultant and TechTarget contributor. "Oracle is notorious for providing confusing information about its products," Sheldon said.

Oracle Cloud Database Service includes two service levels: one that requires users to install the software from VM images and manage it themselves, and another that automatically installs the software and creates databases according to specifications set by users. The latter also includes tools to help automate database backup, recovery, patching and upgrade processes, according to Oracle.

The cloud service supports Oracle Database 11g or 12c, although the 12c Release 2 software that Oracle made available in late 2016 currently isn't available with the virtual-image version. In addition to the main Oracle Database Cloud Service and the bare-metal configuration, the company offers the database software running on its Exadata hardware platform in the cloud, as well as an Exadata Express implementation for small and midsize databases. Users of the cloud service can either pay a monthly fee or an hourly usage rate.

Why not AI?

Topics on the CloudWorld agenda also included adaptive analytics for Oracle's suites of business applications, among other cloud perks. Like other cloud contenders, Oracle has begun to focus on machine learning and AI application enhancements. Its early forays take the form of Adaptive Intelligent Apps, which were initially discussed at the Oracle OpenWorld 2016 event in San Francisco.

This software, accessible via Oracle cloud APIs, embeds analytics and AI capabilities in Oracle's customer relationship management, HR and other applications. Adaptive Intelligent Apps can tap into the company's Data Cloud, which is a large bank of third-party data available as a service. Much of that data was obtained with Oracle's 2014 purchase of Bluekai.

"Oracle has revealed that they're working on an automated machine learning capability that will create models unique to each and every AI Apps customer," said Doug Henschen, an analyst at Constellation Research Inc. "They're also promising 'supervisory controls' that will give customers the ability to tune the machine-learning-based predictions and recommendations."

There are not enough data scientists to meet organizations' needs to automate predictions and recommendations, Henschen said, so AI and machine learning on the cloud is an important step. Oracle sees Adaptive Intelligent Apps as a cloud differentiator, but more detail is needed, he said. He expects the company to launch the first Adaptive Intelligent App in May, with some more applications to become available this summer.

Because machine learning applications rely on large amounts of data, Oracle's connecting its extensive data-as-a-service capabilities becomes a sort of "ace in the hole" in the cloud competition, according to Denis Pombriant, an author and managing principal at Beagle Research Group.

Oracle Data Cloud has billions of clean records, Pombriant said. The records "can be used to augment whatever a business has online, and that leads to better offers and better decisions," he added.

Oracle's overall cloud position is unique, according to Pombriant. Within its large installed base are many large installations, and the migration to the cloud will take years in many of those cases. The company's recent emphasis on IaaS is aimed at customers whose first move to the cloud would be to shift existing applications with little change, he noted.

More availability on the way

Oracle also said it plans to buttress its overall cloud offering by opening three new data centers, or cloud availability regions, over the next six months. These are set for Virginia, the U.K. and Turkey.

An Oracle spokesperson acknowledged that, while the new regional data centers should improve performance, the need to store data locally to comply with privacy requirements in European countries played a role in choosing the new locations.

The company also said it would launch a series of developer events to educate programmers on the open source capabilities of its platform as a service technology. The focus will be on cloud-native development that goes beyond the use of Java to embrace Python, JavaScript and other languages that have become popular in cloud implementations.

Next Steps

Find out why Oracle bought Dyn

Learn about Oracle's cloud-first strategy  

Watch a video on Oracle's cloud database services portfolio

Dig Deeper on Oracle cloud computing infrastructure