Bringing Oracle Cloud Infrastructure On-Prem
Oracle understands how to navigate through the complexity of a multi-cloud world. The company, after all, has spent decades managing the most critical data for nearly every enterprise in the Fortune 1000. By almost any metric, Oracle Cloud Infrastructure (OCI) is the fastest-growing cloud service in the world.
Background: OCI Dedicated Region
Delivering an on-prem solution that provides the benefits of the public cloud model isn’t difficult. Every public cloud provider, and even traditional hardware vendors like Hewlett Packard Enterprise and Dell Technologies that have no public cloud of their own, have on-prem consumption-based services that offer cloud-like pricing flexibility, using the magic of creative financing.
What none of those offerings provide, however, is a seamless experience that crosses cloud boundaries, making the management of on-prem and cloud indistinguishable. Oracle is the only company that provides that today, greatly simplifying the experience for enterprise IT.
Two years ago, Oracle introduced OCI Dedicated Region to extend Oracle’s cloud infrastructure to on-prem data centers. An OCI Dedicated Region looks exactly like an OCI public cloud region but delivered as an independent, single-tenant platform hosted in the customer’s data center. OCI Dedicated Regions are built, owned, and fully managed by Oracle and supported as a service.
Background: Oracle Cloud@Customer
While OCI Dedicated Regions bring tremendous value to the enterprise, it’s not a lightweight solution. OCI Dedicated Regions scale from 12 racks to more than 450. That’s a big footprint, one that comes with minimum purchase requirements. Oracle simplifies on-prem deployments even more for customers for whom a Dedicated Region might be too large with its Cloud@Customer model that start with a single rack.
Oracle Cloud@Customer brings a targeted set of capabilities into the data center. It does this with the same seamless integration and management experience with OCI that a Dedicated Region provides. Oracle first introduced its Cloud@Customer model with the Exadata Cloud@Customer offering.
Exadata Cloud@Customer allows enterprises to leverage the power of Oracle’s high-performance Exadata Database Machine, but with a cloud consumption and management model. Exadata Cloud@Customer isn’t just another on-prem consumption-based model; it’s fully integrated and managed through the OCI control plane. Customers can run Exadata Database Service and Autonomous Database Service simultaneously on Exadata Cloud@Customer
News: Introducing OCI Compute Cloud@Customer
Oracle this week extended its Cloud@Customer model by introducing its new Compute Cloud@Customer offering. The new service delivers the same compute infrastructure deployed within OCI public cloud, including compute, storage, and networking capabilities, in a rack-level on-prem solution.
Oracle Compute Cloud@Customer enables enterprises to migrate and run Oracle and third-party applications and develop cloud-native applications that can be seamlessly run and managed across Oracle’s distributed cloud, including on-prem and public cloud, without modification. This allows organizations to modernize their application stacks and deploy at scale while allowing workloads to run in the most appropriate location, whether for compliance or performance reasons. That’s powerful.
Oracle’s new Compute Cloud@Customer offers several key capabilities:
- Unified user identity management (IAM): unified identity management across OCI public cloud regions, OCI Dedicated Region, and Compute Cloud@Customer.
- Portable Infrastructure-as-Code: OCI IaaS APIs for full-portable VM images that run on identical compute, storage, and networking.
- Secure Operator Access Control: enables secure, granular, and auditable access spanning OCI public cloud regions, Dedicated Cloud, and Compute Cloud@Customer.
- Seamless updates and patches: customer-defined, automated updates delivered from OCI public cloud regions and Dedicated Regions.
- Exadata Cloud@Customer Direct Connect: low-latency connection from the application tier running on Compute Cloud@Customer to the database tier, accelerating application response times.
Oracle offers Compute Cloud@Customer in four primary flavors, with multiple expansion and upgrade offerings. Based on AMD’s latest generation EPYC processor, Oracle’s baseline offering delivers 552 processor cores, 6.7 TB of memory, and 150 TB of usable storage. Compute and storage scale independently to 6,624 cores and 3.4 PB of storage while memory scales to 80.4 TB at the high-end. There are multiple storage service options—file, block, and object which can be mixed and changed on demand. No other provider offers this set of choices.
Pricing for Computer Cloud@Customer is the same as Oracle’s existing OCI pricing. The offering is sold as an infrastructure subscription with monthly fixed costs coupled with a metered consumption pay-only-for-what-you-use model for OCI IaaS services such as compute, memory, storage, and networking. It’s a very compelling offering.
Analysis
The public cloud providers have changed how we look at the enterprise infrastructure business. IT buyers like the consumption-based models pioneered by the cloud providers. Unfortunately, IT administrators don’t like the struggle that seems to follow most hybrid-cloud approaches. Complexity and costs are the enemy of digital transformation projects.
IT buyers like the operational flexibility that the public cloud provides. Capital expenditures go away. OpEx becomes predictable. Resources are available on-demand, with near-endless scalability. There’s a lot to like, but the off-site nature of the public cloud often complicates an enterprise’s ability to fully leverage that flexibility.
Despite all its benefits, the public cloud still can’t compete with traditional on-prem IT when it comes to matters like performance, security, and data governance. Nearly every enterprise has data that simply can’t leave the building and latency-sensitive applications. Splitting infrastructure to run across cloud boundaries in an attempt to have the best of both worlds introduces costs and complexity that can stall a digital transformation effort.
Oracle has solved the problem of hybrid-cloud infrastructure. With OCI, Oracle has built one of the market’s most robust and fastest-growing public-cloud businesses. The company has gone even further, solving the complexity of on-prem deployment with an innovative model that seamlessly blends public cloud regions, Dedicated Regions, and Cloud@Customer.
Oracle isn’t the only cloud provider offering on-prem compute services. AWS has Outposts, Azure has its Azure Stack Hub, and Google has its Google Distributed Cloud Edge. What distinguishes Oracle is that Oracle Compute Cloud@Customer is a completely integrated cloud platform—built, installed, owned, and managed by Oracle—that looks exactly like its OCI public cloud infrastructure.
Oracle provides a transparent experience, with consistent hardware, software, and control plane allowing for seamless interoperability across OCI public cloud regions, Dedicated Regions, and, now, Oracle Cloud@Customer. It is entirely compatible with the OCI public cloud regions and OCI Dedicated Region and delivers the same core OCI cloud services, at the same low pricing.
Oracle’s competitive differentiation extends beyond just transparent interoperability. Oracle provides a more scalable experience. Oracle Compute Cloud@Customer scales to 2,208 processor cores per rack and 96 cores per virtual machine (VM). Azure comes closest to matching Oracle’s capabilities, providing 1,536 cores per rack and 32 cores per VM, but that’s still a significant deficit for Azure. AWS and Google compare even worse. The same is true when looking at storage capabilities.
Oracle’s new Compute Cloud@Customer brings Oracle Cloud Infrastructure on-prem in an accessible and affordable way that is fully compatible across Oracle’s distributed cloud. Any enterprise deploying a hybrid-cloud solution will not find a solution that surpasses the one Oracle delivers.
