By Margaret Harrist
Larry Ellison unveils new Oracle Cloud Platform services.
On June 22, Oracle announced that it is not only extending its big data technology to the cloud, it’s giving organizations an easy way to ramp up and secure a big data project quickly—and to avoid “noisy neighbor” performance issues in the cloud.
Oracle Big Data Cloud Service and Oracle Big Data SQL Cloud Service were among more than 20 new Oracle Cloud Platform offerings announced by Oracle Executive Chairman and CTO Larry Ellison during a launch event at company headquarters in Redwood Shores, California.
“We’ve entered a new era of cloud computing,” Ellison said, emphasizing that Oracle is the only cloud services provider playing at all three layers of the cloud—software as a service (SaaS), platform as a service (PaaS), and infrastructure as a service (IaaS).
As big data rapidly becomes a vital part of enterprise information architectures, organizations want to increase their big data capabilities and make it easier to explore new opportunities. Oracle Big Data Cloud Service is designed to meet those needs, and to deliver consistent high performance.
With Oracle Big Data Cloud Service, organizations can build a discovery lab to validate tools or test a methodology with low risk and little investment. IT departments can get the latest tools and technology with the click of a button—and then shut the lab down quickly if the project doesn’t work out. The new service also bridges the needs of developers, who want the latest and greatest technologies, and IT people, who want stable systems and simple management.
Taking cloud-based big data capabilities a step further, Oracle Big Data SQL Cloud Service enables organizations that use both Oracle Big Data Cloud Service and Oracle Database Cloud Exadata Service to use simple SQL queries to access data across their Hadoop, NoSQL, and relational data stores. This service also extends Oracle Database security across Hadoop and minimizes data movement, which increases performance significantly.
You can push a button and move an application from on-premises into the cloud. And by the way, you can push a button and move a database from the cloud back on premises.
—Larry Ellison, Oracle Chairman of the Board and CTO
If you’ve ever lived in an apartment with thin walls, you’re painfully aware of how disturbing noisy neighbors can be.
Your organization can experience a similar hassle in the cloud—if the cloud environment is structured so that you share underlying CPUs, disks, and a network with other organizations. With some cloud providers, you may pay for a virtual CPU core, but the physical CPU that backs it may be shared among many organizations.
“If two organizations are sharing CPUs, disks, or networks in the cloud and one of them is generating a lot of network traffic, their neighbors can experience a lot of variability on their performance,” explains Dan McClary, senior principal product manager, Oracle Big Data. “Cloud administrators might decide to move that ‘noisy neighbor’ organization to another location so that its neighbors’ performance won’t suffer. But getting moved will result in a change in performance—and if that happens when you have a lot of demand, your performance will take a hit at a really bad time.”
Sometimes, the “noisy neighbor” organization may not even realize it’s being noisy, so being moved will seem arbitrary. “Big data applications involve a lot of network traffic, so you could be running a normal big data job and it could be flagged as noisy. You’d get moved, and your performance and your SLAs would suffer,” McClary says.
With Oracle Big Data Cloud Service, organizations have no cloud neighbors; the instances and network are dedicated to their workload.
“Unlike many other clouds, when you set up a big data cloud environment on Oracle Cloud, you get consistently high performance even if there’s a raging party down the block—or if you’re the one having the party,” McClary says.
For organizations considering the DIY Hadoop route, Oracle Big Data Cloud Service provides an alternative that’s easy to set up and far easier to manage. Rather than installing updates and patches and reconfiguring systems, Oracle Big Data Cloud Service delivers the latest innovations in the Hadoop ecosystem as well as unique innovations from Oracle that will enable organizations to speed their ability to perform big data analytics, including:
“All of this is automatically set up and configured out of the box, no extra charge,” McClary says. “There’s no need to install a Cloudera distribution or pay additional for Cloudera support.”
And any tools that are certified on Oracle Big Data Appliance or on Cloudera are certified on Oracle Big Data Cloud Service—like SAS High-Performance Analytics and NGDATA’s Lily.
Accenture’s 2014 report “Big Success with Big Data”, which included survey data from more than 1,000 people at companies that had completed at least one big data project, found that the top challenge with big data projects is security (listed by 51 percent of respondents).
And it’s no wonder: Setting up security in a big data environment is a laborious, multistep process that requires the establishment of a Kerberos network authentication protocol—a complex undertaking—followed by configuring Hadoop instances to Kerberos, setting up an LDAP, configuring Apache, and setting up encryption. With Oracle Big Data Cloud Service, you skip all those steps.
“Security in Hadoop solutions overall has improved, but getting it up and going is still a hassle,” McClary says. “With the Oracle solution, you just check a single box on the Oracle website when you set up your cloud service, and we automate all the setup and configuration of the security.”
At the launch event, Ellison described a range of new cloud services—including Oracle Storage Cloud Service, Oracle Mobile Cloud Service, and Oracle Process Cloud Service. “We have a complete suite of services for the creation of modern applications and the running and managing of those applications,” Ellison said. “So our PaaS and IaaS are complete. I’m not saying we’re not going to keep improving everything. I’m just saying all the major boxes are filled in.”
One of Oracle’s biggest differentiators is the ability to move databases and applications from on premises to the cloud—and back again—quickly and easily. This is key to helping customers work through the coming decade of coexistence between cloud data processing and on-premises data processing. “You can move applications back and forth, move data back and forth,” Ellison said. “Everything is not going to move to the cloud by this Friday. People are still going to need to manage both on-premises and cloud. We make that easy.”