Oracle expanded the capabilities of its Exadata Cloud at Customer hardware service with a multi-VM autonomous database feature that became generally available today.
Oracle Cloud at Customer enables organizations to use Oracle Cloud Infrastructure in an on-premises deployment.
The approach is similar to the one other major cloud providers are using, such as Amazon with its AWS Outposts platform.
Autonomous and non-autonomous
The Exadata platform is Oracle's on-premises hardware for running the Oracle Database. Until now, Exadata hasn't been running the Oracle Autonomous Database -- which is a fully managed service -- but rather the non-autonomous version of Oracle database.
With the new Oracle Multi-VM Autonomous Database capabilities for the Exadata Cloud at Customer, organizations can now run both types of Oracle database (autonomous and non-autonomous) on the same physical infrastructure.
Ron WestfallAnalyst, Futurum
With this move, Oracle is now taking on-premises database automation to a new level by providing a full self-service database cloud option, said Ron Westfall, an analyst at Futurum Research.
Oracle Multi-VM Autonomous Database on Exadata Cloud at Customer delivers cloud database functionality in customer data center deployments that can meet the needs of Oracle's database users, Westfall said.
"[Exadata Cloud at Customer] provides the cloud database framework needed to ease data management throughout database environments and accelerate transitions to the cloud by enabling existing applications to simply connect and run," he said.
How Multi-VM Autonomous Database works
It's important for Oracle to enable its users to run multiple versions of the Oracle database.
Robert Greene, vice president product management/strategy at Oracle, noted that many organizations have large database deployments that can run hundreds or even thousands of Oracle database instances.
With such large deployments, Greene said it can be a significant challenge to move existing on-premises Oracle databases to the Oracle Autonomous Database. Meanwhile, many organizations will need to support and operate the non-autonomous Oracle database deployments for years to come, he added.
It was possible for organizations to run both types of Oracle database before the release of Multi-VM Autonomous Database for Exadata, but Greene said it would have been somewhat more costly with users running the two type databases on different hardware or cloud services.
"In the past it would take a fairly substantial second capital investment in order to run both of these non-autonomous and autonomous systems," Greene said. "Now, they can do this all together. Through virtualization, they can deploy multiple virtual machine clusters onto Exadata systems."
Greene explained that the virtualization technology that Oracle is using to enable the new feature is based on Kernel-based Virtual Machine (KVM) hypervisor clustering.
KVM is part of Oracle Linux, the base operating system that the Exadata system runs. The Multi-VM Autonomous Database inside of the Exadata hardware also uses Oracle Real Application Clusters and integrates the RDMA over Converged Ethernet protocol that helps enable 100 Gbps network connectivity across virtual machines.
Providing support for Multi-VM Autonomous Database is a somewhat belated move for Oracle.
Greene acknowledged that it's a capability that Oracle customers have been asking about for quite a long time. But it was a difficult and involved effort to deliver the service in a way that Oracle could guarantee service-level agreements across mixed-mode environments, he said.
"It was just an engineering effort for us to get it done," Green said. "We've been working on it for quite a long time, and it's now ready."