Not all databases or data centres are made equal, according to Steve Daheb, senior vice-president of Oracle Cloud.
Speaking to the media during the Oracle OpenWorld conference in London this week, Dehab outlined the advantages of his company's Autonomous Database, which, he said, uses machine learning to optimise and simplify cloud database management.
Question: Some people say Oracle is trying to fence its customers in with its cloud version 2 strategy.
Daheb: With Gen 2, we mean multiple things. We mean the new infrastructure we've rolled out is based on bare metal (installed directly on hardware) that will support AI [artificial intelligence] and machine learning. A big part of Gen 2 is our Autonomous Database, which has machine learning integrated into the database. For us, it's about continuing to innovate. We have thousands of trials now with our customers who are trying the Autonomous Database and they're seeing the benefits of it.
Question: Please talk about the lock in.
Daheb: You can run Oracle Database on-premises and on different hardware. You can run it on Amazon and Azure. And you can run Oracle Database on Oracle. Where else can you run Amazon's [data warehouse service] Redshift? Only on Amazon. Where else can you run [the relational database service] Aurora? Only on Amazon.
To be honest, you can run Oracle Database in more places than you can Amazon's own [services]. In terms of Autonomous, it's different, because there's a secret sauce to Autonomous. It's thousands of engineering years' worth of database optimisation. Autonomous is how we've integrated AI and machine learning into our database. Foundational to that is our infrastructure. You couldn't give somebody our Autonomous code and get the same results.
Question: Where are you in terms of rolling out your infrastructure, compared to the other big cloud giants?
Daheb: There's not a single customer who needs Autonomous Database where we would not be able to provide it. We continue to expand our data centre presence.
Question: I'm in Johannesburg. Where will I be served from?
Daheb: There are multiple locations. But you're not going to get the latency. If it's a data sovereignty issue, we can also utilise the 'Cloud at Customer' solution, whereby we can deploy an Exadata Cloud Machine (that runs Oracle databases on-premises) behind the customer's firewalls.
I think Amazon just announced something similar to what we've been delivering for two years with respect to data privacy. There's many different ways for us to meet those customer needs. If somebody wants Autonomous Database capabilities, whether it's on-premises, or Cloud at Customer, or whether it's within our cloud, we'll be able to meet their needs. We continue to expand, and open up more data centres. We're in a good position.
Question: Even if Oracle triples or quadruples the amount of data centres, in 2018, your investment in data centres has been slowing in terms of absolute figures, related to revenue, as well as the percentage. Does this mean you don't have enough money to compete with the big players in terms of infrastructure-as-a-service?
Daheb: I think we have a thoughtful roadmap with respect to infrastructure that will meet the needs and demands of our customers. If I'm deploying an Exadata in the same square footage as a different cloud provider that's using X86 or off the shelf components, I guarantee that we're going to be much more optimised per square foot.
People need to start thinking about it differently. It's not just about the size of your data centre. I can have a car engine, but if I fit a V4 versus a V8 in the same space, I'm going to be able to do much more. Exadatas are highly optimised and support multi-tenancy and dense storage. We're one of the biggest cloud application vendors in the world and that all runs in data centres that we have on a global basis.
I think it's an old paradigm to think about square footage and the amount of the data centres versus when you integrate machine learning technology, when you have Exadata Database technology. There's a lot more you can offer. I would urge people, when they look at our offering, to look at it holistically. It's a very complete cloud.
Question: You're saying things run better on Oracle. Why is that?
Daheb: Database definitely runs better on Oracle. After 40 years of inventing and optimising the database, we have people who understand how to optimise performance and queries; how to make technology like machine learning and apply it to query performance. We can actually test a query to make sure it has optimum performance before it implements it. There's amazing things going on. Look at the data. We have customers who say 'I've moved off AWS and moved to Oracle, and I'm seeing an 8X performance increase'. I didn't make that up. It's what they're telling us.
Share