Larry Ellison, Oracle’s cofounder, executive chairman, and chief technology officer, has said Oracle’s new autonomous database ranks among “the most important things we’ve ever done.” That’s because it deploys, optimizes, patches, and secures itself with no human intervention, bringing new levels of performance, security, and efficiency.
So how is this autonomous database playing in the real world? Oracle Autonomous Data Warehouse Cloud Service has been available since March. I talked with five database users at this year’s Collaborate 18 about their reaction to the technology and where they see this broader autonomous computing push going:
Jim Czuprynski: What Do You Need DBAs For Now?
After plying the trade as a database administrator for 20 years, Jim Czuprynski considers himself a DBA’s DBA. “I know the daily stresses and joys of being a DBA,” he says.
So Czuprynski, an enterprise architect at Vion Corporation, and a technology blogger, admits that when he first encountered autonomous database technology, his knee-jerk reaction was to ask, “If a computer can do what I can do, then what do you need me for?” he says.
It didn’t take him long to change his tune. “Oracle is simply taking what they’ve been telling us for years are best practices and having the machines just do it,” he says. The autonomous machines “never get tired and can watch everything all at once.”
For example, he says, “the machines are going to watch your database perform against its application workload over time and then it’s going to react to it, subtly, with machine learning,” to tune and better secure the database. “Why wouldn’t I go for that?” he now asks. “Why wouldn’t I want to just turn that over to something that’s smarter than me?”
That still leaves plenty for Czuprynski to do. “It doesn’t mean I don’t have to think out the data model, it doesn’t mean I don’t have to write better application code against the database” to make the applications faster or more reliable, he says.
In the end, autonomous database technology is about giving time back to data experts to do more important work. “You’re going to have a lot more time to spend with developers and with the people that are running the business,” Czuprynski concludes. And that, he says, is a good thing for a DBAs career.
Nitin Vengurlekar: Autonomous Is About Business Agility
Nitin Vengurlekar is CTO at Viscosity, a firm that helps his clients get the most out of their data. “To me, autonomous database equates to a faster speed of business,” he says.
Vengurlekar likes cloud services in general because they provide infrastructure on demand, “so I don’t have to wait weeks or months for technicians to set up the environment,” he says.
Autonomous database technology turns that speed up a notch higher, by taking the database tuning and management off his plate while giving him the three things he really wants in technology for his business: “I want it easy, fast, and elastic,” he says.
Vengurlekar gets easy and fast with the Oracle Autonomous Data Warehouse Cloud Service because he can “work with data right away without having to assign DBAs to make sure it’s set up,” he says.
“And I want to grow and shrink the data warehouse based on what my needs are,” whether that’s end of the year or month, or different holiday seasons, he says that Oracle Autonomous Data Warehouse Cloud Service "does that also really well.”
For speed, Vengurlekar did his own benchmarks to understand how Oracle Autonomous Data Warehouse compares to other cloud-based data warehouses: “How many queries can I run in an hour? If I have to pay for the service for an hour, how much can I churn through with it?” He found that Oracle delivers on its claim to deliver queries many times faster than other data warehouses he tried.
The speed means, “I can do faster analytics on diverse data sets and have power for things like faster visual analytics and predictive pattern management,” he says.
Russ Lowenthal: Patching Is the Killer App
An autonomous database is worth it for patching alone, says Russ Lowenthal, a director of database security at Oracle who hears from many database users on the subject.
“If I’m a CIO, my number one concern is that a data breach like those in the news lately will happen at my company. That means I’ve got a system out there, and it’s got a known vulnerability and a patch exists for it, but I haven’t applied the patch to it yet,” he says. “That’s my nightmare.”
An autonomous database from Oracle patches itself as soon as a vulnerability is detected and a patch is ready, without taking the system offline. “That immediately takes a huge burden off of the shoulders of CIOs and DBAs,” says Lowenthal.
Patching is hard. To make the case, Lowenthal gives an extreme—but very real—example: “I’ve got one customer with 17,000 Oracle databases. We release patches four times a year. If they patch each database four times per year and each patch takes an hour to apply, that’s 35 full-time workers doing nothing but patching.”
What CIO wouldn’t want to put those thirty-five people “on cool stuff like data engineering or data mining and other valuable, business-centric functions?”
Shrinking the human touch points and thus human hours needed for manual tasks like patching goes a long way toward improving security and reducing error. Plus, he adds, “I don’t know any DBA who enjoys patching.”
Michelle Malcher: Security Functions Deployed Without Effort
Michelle Malcher agrees that patching is a “killer app” of autonomous databases but says there’s a lot more to the security story. Another benefit is that services like the Oracle Autonomous Database and Data Warehouse Service run all available security features of the database by default. “And because there are no humans involved, you avoid exploitable mistakes in the setup,” says Malcher, a data security architect at Extreme Scale Solutions.
Having database options like encryption, secure backups, and database vault turned on automatically is a great start to a more secure environment, Malcher says. The security posture is further improved by managing the database with log monitoring and machine learning in Oracle Cloud.
Once you deploy your database in a secure configuration, the system monitors any changes from the initial install. “It will look for data moves that are at strange times or users that are changing permissions and it will note those,” she says. If it alerts a false positive, for example, learns that behavior as normal. “It learns patterns of behavior over time.”
Dan Vlamis: I Want More Autonomous Technology
Once Dan Vlamis spent time working with Oracle Autonomous Data Warehouse Cloud, he began asking himself why more of his technology isn’t autonomous.
Vlamis’s firm helps companies see into their corporate data by building business intelligence dashboards that are easy on the eyes and tell a good story.
“We found that we could load the client records very quickly and do the analysis without building indexes or partitioning the data or worrying about performance,” Vlamis says of using the autonomous data warehouse cloud service.
“That got me excited about other ways my technology could step in and take over mundane tasks,” he says. For example, he says “when we build a dashboard, we have to figure out what color palette and how many colors we want to use.” He’d like a future autonomous system to figure out what to do by looking at the types of measures and key performance indicators he’s after. “It would know that I need a divergent color scheme or a qualitative color scheme and just say, ‘Here, I’ll just do it for you,’” he says.
Now that he’s used an autonomous cloud service, he’s hooked: “I love anything that frees me up to do other things in my business.”
Jeff Erickson is editor-at-large for Oracle.
