OpenWorld 2017: Ellison declares autonomous database 'revolutionary'

In an ebullient conference keynote that shows Oracle still has Amazon Web Services (AWS) chiefly in its sights, chairman and CTO Larry Ellison declared the firm’s 18c version of its database the “biggest thing we’ve done in decades”.

He told attendees of Oracle OpenWorld 2017 in San Francisco that the database, which he says “runs and tunes itself” deserves the soubriquet “revolutionary”, adding that he does not use the “Silicon Valley” term lightly.

In a speech in which he had, as he did at last year’s show, much sport at Amazon’s expense, he said Amazon Elastic Compute Cloud (EC2) was “anything but elastic”, and that the rival supplier spent in the region of $60m with Oracle in 2016.

As he concluded the first of his two keynotes at the event, he also took a sideswipe at business applications supplier rival SAP. “They acquired three major online services years ago now, and they all run on the Oracle database, not on SAP Hana. SuccessFactors still runs on Oracle. SAP is one of the biggest Oracle users on the planet,” he said.

Ellison linked the announcement of the new version of the database – dubbed Oracle Autonomous Database Cloud, and said to be the world’s first autonomous database – with the salience of cyber security use cases.

With the world’s first 100% self-driving autonomous database, and new automated cyber defense applications that detect and remediate attacks in real time, Ellison added: “Data theft is the biggest cyber security threat. The safest place for data is autonomous Oracle database.”

Ellison said this is because patching by humans – such as what failed to occur in the recent Equifax data loss case – is a structural vulnerability. But, he added, database automation will free up database professionals to do more important things, and will not eliminate database administrators altogether.

He also said the machine learning underlying the new database to be “as revolutionary as the internet itself”, adding that “distinguishing between normal and abnormal patterns in data is the critical application of machine learning”.

The main pitch to attendees, buttressed by an extended comparison of how Oracle runs against Amazon, was that his company’s technology runs “five to eight times faster than Amazon”, and that “labour costs are almost gone”.

“We will write in contracts that your [database] bill will be half in the Oracle cloud compared with [data warehouse service] Amazon RedShift,” said Ellison.

In a statement, Oracle said: “With total automation based on machine learning, Oracle Autonomous Database Cloud eliminates the human labour required to manage a database by enabling a database to automatically upgrade, patch and tune itself while running.

“With no more scope for human error or requirements for human performance testing, Oracle is able to minimise costly planned and unplanned downtime to less than 30 minutes a year and guarantee that organisations can cut their costs in half compared to Amazon.”