In the first of his two scheduled keynotes at Oracle’s OpenWorld conference, Oracle’s CTO and executive chairman Larry Ellison detailed the benefits of the company’s Autonomous Database, and touched on the Highly Automated Cyber Security offerings.
Oracle’s new technologies are both powered by machine learning, a form of artificial intelligence which enables computers to learn from data patterns to make predictions. Machine learning can be applied to detect anomalies in security scenarios, Ellison said. It can determine for example if a CFO based in California has logged into a secure system from the Ukraine, or if it is an attempted breach.
“The cybersecurity technology is designed to automatically detect threats when they first occur, and then direct the database to immediately remediate that problem,” Ellison told the attendees gathered at the Moscone Center. “People are going to get better at stealing data, so we’ve got to get better at protecting it.”
Oracle’s Autonomous Database, version 18C, is “the world’s first and only” truly autonomous database, said Ellison. And it is “the most important thing we have done in a long, long time.” The database, he said, “is 100% automated.” It “automatically provisions itself, patches itself, updates itself, tunes itself, without any human intervention whatsoever.”
Since the technology can patch itself while running, it can significantly reduce the potential complications that take place when humans manually schedule their own downtimes. According to Ellison, “the worst data thefts in history occurred after a patch was available to prevent theft. The patches just weren’t applied.” Relying on humans, he stressed, results in errors. The process many companies typically employ with the tools available to them “doesn’t work.” He pointed out Equifax as an example of a recent failure. “We have to automate our cyber defenses,” and to be able to defend databases without having to take them offline or shut them down, he said.
The Autonomous Database can adapt and tune itself continuously, as query patterns and data distributions change, and as the sizes of databases change. “This is a big deal, by the way; no one else does this,” Ellison said.
In its Service Level Agreement, Oracle promises “zero” labor costs for systems administration and management. “When you eliminate human labor, you eliminate human error,” as well opportunities for malicious behavior from within an organization, Ellison said. It also enables database professionals to focus more of their time on important issues such as design, analytics, policies and authorizations.
Because there are no humans involved, also, the company can guarantee 99.995% reliability and availability. The systems will experience planned or unplanned down time of “less than thirty minutes a year with the Oracle Autonomous Database,” Ellison said. This contrasts with Amazon’s Redshift offering, which is entirely manual, and has far more downtime.
Oracle’s Autonomous Database is elastic in its use of computer resources, meaning that if a company needs more compute capacity to run a big query, it can switch from eight processors to 16 processors, for instance, and drop back to eight when possible. This is also in contrast with Amazon’s Elastic Compute Cloud, which “just isn’t elastic,” Ellison said.
Ellison promised that users will see huge savings compared to those on Amazon, repeating that they must to be “willing to pay much less.” “All of this stuff costs a fraction of what Redshift costs at Amazon, or a fraction of what other databases cost in other clouds,” he said. He promised that the bill would be half as big as it would be with Amazon. “We will write that in your contract.”
The Oracle Autonomous Database runs on Exadata infrastructure. it’s available in Oracle’s public cloud, on premises, and through its Cloud at Customer offering.
The Autonomous Database for Data Warehouses will be available in December of 2017, and an online transaction processing (OLTP) version will be available in June of 2018.