Cloud Computing

Pick Your Horse Wisely

From cloud to big data and in-memory analytics, which database technology will you ride through the transitions ahead?

by Jeff Erickson

Fall 2016

The database world is going through three simultaneous upheavals: cloud computing, big data, and in-memory analytics. Each offers business opportunities for those who can seize them—and trouble for those who can’t. Andy Mendelsohn, executive vice president of database server technologies at Oracle, is responsible for making sure Oracle Database is way ahead on all three fronts. He shared with Profit some of the latest innovations in Oracle Database 12c Release 2 and Oracle’s strategy for making sure customers come out on top.

Profit: With all the databases popping up in the cloud, why are customers choosing Oracle Database?


Mendelsohn: Given our long track record of database innovation, customers trust that Oracle is going to continue to innovate to ensure it remains the #1 database as they transition to the era of cloud, big data, and in-memory analytics. Another reason is that as we add new innovations such as the Oracle Database In-Memory and Oracle Multitenant options for Oracle Database 12c, we always ensure that existing applications can benefit without requiring application changes.

Customers can take all those applications they’ve written in-house for Oracle Database—all their independent software vendor [ISV] applications—and just run them on a cloud using Oracle Database without any change to the applications. Customers can leverage all their developer and administrator skills and get the same support for their on-premises and cloud databases.

Profit: What about Oracle Database in Oracle Cloud?

Mendelsohn: Database customers will find that all clouds are not equal. Only Oracle Cloud offers infrastructure optimized for Oracle Database. So all critical Oracle features such as Oracle Real Application Clusters, Oracle Active Data Guard, and scale-out parallel query run well out of the box on Oracle Cloud but are generally not available on other clouds such as Amazon Web Services.

Profit: How does Oracle Database 12c make the cloud better?

Database customers will find that all clouds are not equal.”–Andy Mendelsohn, Executive Vice President of Database Server Technologies, Oracle

Mendelsohn: In Oracle Database 12c, we rearchitected the database to help customers take advantage of the cloud more easily. Using Oracle Multitenant, administrators can manage many pluggable databases as one in a single multi-tenant container database. They can back up and recover, upgrade, and patch up to thousands of pluggable databases inside a single container, in a single operation.

It also makes the databases much more mobile. Administrators can unplug a pluggable database from one database container and plug it into another one. That could be moving a pluggable database from one data center to another or—in a hybrid cloud environment—it could be moving a pluggable database from on premises to Oracle Cloud, or vice versa.

Profit: What about big data?

Mendelsohn: Businesses often maintain multiple systems for big data analytics—for example, they might have a big data warehouse for analyzing transactional data from enterprise applications and another system running Hadoop or NoSQL for holding web and Internet of Things (IoT) data. For Oracle Database 12c, we brought these data management systems together with a technology called Oracle Big Data SQL. Now you can run standard Oracle Database SQL in a massively parallel way across your Oracle, Hadoop, and NoSQL databases. You don’t need to move data between systems.

Profit: How do you plan to stay on top of in-memory analytics?

Mendelsohn: In Oracle Database 12c Release 1 [12.1], we introduced our revolutionary column store technology for in-memory analytics. This technology is unique in that we let you do high-performance analytics against live transactional data all in the same database. Our competitors don’t let you do that. They make you copy the data to different tables or a different database to do analytics.

The most frequently requested enhancement to the version 12.1 in-memory column store technology was the ability to isolate the analytics users from the transaction users. In Oracle Database 12.2, we addressed this with Oracle Active Data Guard: transaction users run on the primary database and in-memory column stores are created for the analytics users only on the standby database.

While we deal with the technology changes in front of us, we need to prepare for what’s on the horizon as well. The impending arrival of persistent memory or nonvolatile DRAM will allow systems to leave the entire database in memory at all times. The whole architecture of databases is going to be completely disrupted by this change. We are putting big investments into Oracle Database to support this next transition.

Action Items
  • SEI Speeds DevTest and Time to Market with Oracle PaaS
  • Find more new ideas for cloud-driven transformation
  • Photography by Shutterstock