New cloud partnerships put your data where you need it

The multicloud database era has arrived, opening exciting new possibilities for how businesses manage their data in and across clouds. Upsides include greater choice in platforms and cloud service providers, more flexibility in workloads and data distribution, and split-second responsiveness.

And the timing couldn’t be better: Development teams working on artificial intelligence projects will now find it much easier to build applications that blend data from their Oracle Databases with AI tools and services from other leading hyperscalers.

Three major developments to advance multicloud databases came together in September at Oracle CloudWorld 2024:

1. Oracle Chairman Larry Ellison disclosed a groundbreaking agreement with AWS to host Oracle Databases running on Oracle Cloud Infrastructure (OCI) in AWS data centers. The new Oracle Database@AWS service creates native integrations between Oracle Autonomous Database or Oracle Exadata Database Service and AWS services. This means, for example, that customers can connect their Oracle Databases to AWS services, such as Amazon Elastic Compute Cloud (EC2) and Amazon Bedrock for generative AI development.

2. In a similar partnership with Microsoft, Oracle expanded the availability of Oracle Database@Azure to more global regions, now reaching Asia and South America. With this service, first introduced in 2022, customers can provision, access, and monitor Oracle Database services running on OCI from within the Azure cloud.

3. And finally, Oracle Database@Google Cloud, announced in June, became generally available in four regions in the United States and Europe, with additional regions planned. Prior to this, Google Cloud made its Cross-Cloud Interconnect available in 11 global regions, enabling customers to move data and workloads without data transfer fees. Now, with Oracle Database@Google Cloud, customers can purchase Oracle Database in Google Cloud Marketplace, and integrate their databases with Google services, such as Gemini foundation models, Vertex AI, BigQuery, and Looker.

Each of these multicloud partnerships is noteworthy on its own. Collectively, they represent a tectonic shift in the database landscape—and not only because longtime rivals have set aside competitive interests to give customers more options with less complexity. These multicloud collaborations have the potential to accelerate all kinds of digital initiatives, including GenAI development, analytics, database consolidation and migrations to the cloud, and data distribution across regions for resiliency or to help meet data residency requirements.

“Customers wanted us to find a way to make the very latest and best Oracle technology available on other clouds in addition to OCI. We found a way,” Ellison said on Oracle’s FY25 Q1 earnings call. “This will enable customers to use the Oracle Database anywhere and everywhere.”

Many of these new capabilities are the result of native integrations, which are pre-engineered integrations of multivendor software and services, so they look, feel, and work as one. In this column, I’m going to dig into how that works and the many advantages that come with it.

From interconnects to collocation

These multicloud advances were jump-started in 2019 when Oracle partnered with Microsoft to launch Oracle Interconnect for Microsoft Azure. That and a similar deal with Google Cloud were the first steps to build low-latency, high-throughput direct connections between OCI, on the one hand, and Microsoft Azure and Google Cloud, respectively.

These interconnects enable fast Oracle Database query performance, with latency in the range of 2 milliseconds, and bring collaborative support by Oracle and its cloud partners as well as unified identity and access management. Typical use cases include running enterprise apps across clouds or establishing interoperability between complementary services, such as Oracle Cloud Infrastructure Kubernetes Engine and Google Spanner.

Oracle and its partners are taking their technical integrations even deeper with the Oracle Database@ cloud services. These go beyond network interconnects by colocating Oracle Exadata and Oracle Cloud Infrastructure (OCI) within Oracle’s cloud partners’ data centers. This can accelerate database response—from milliseconds to microseconds in some cases. That’s a requirement for latency-sensitive workloads, such as AI, analytics, and transaction processing.

The business advantages are compelling. Ellison talked about this new phase of open, multicloud services in public conversations with Microsoft CEO Satya Nadella, AWS CEO Matt Garman, and Alphabet CEO Sundar Pichai.

What’s everyone so excited about?

I see the value coming first and foremost in four areas: ease of deployment, application performance, data management, and a richer environment for AI development. Let’s look at each.

Ease of deployment, administration, and use

It hasn’t always been easy to deploy and manage Oracle database management systems in non-Oracle clouds. Of course, there are ways to do it, such as running a self-managed Oracle Database on bare metal servers or via third-party services, such as Amazon RDS for Oracle, but these come with trade-offs, including less functionality and more complexity.

One way to avoid compromises is to run Oracle Database on OCI, which ensures full functionality along with price and performance advantages, and that continues to be an excellent way to go. But some organizations that have applications on AWS, Azure, or Google Cloud want to use data stored in Oracle Databases. That’s now easier than ever. For example, with Oracle Database@AWS, customers can purchase Oracle Exadata Database Service or Oracle Autonomous Database directly from the AWS Marketplace, then configure and launch using familiar tools, such as the AWS Management Console. Similar processes are in place for Azure and Google Cloud customers.

With this one-stop shopping comes simplified database administration and unified support, which is music to the ears of DBAs. And a no-fuss purchasing and billing process lets customers apply existing licenses, benefits, commitments, and discounts, so you can get started quickly.

Customers can also access, or soon will be able to, other enterprise-class Oracle services, such as Oracle Database Zero Data Loss Autonomous Recovery Service, OCI GoldenGate, and Oracle Data Safe, through these same marketplaces.

From millisecond to microsecond performance

One of the major advantages of this new multicloud model is that performance is on par with what Oracle customers already get with Oracle Database on OCI. Because Oracle Exadata and OCI services are colocated within the AWS, Azure, and Google Cloud data centers, there’s no dip in performance from what customers who use Exadata or OCI are already accustomed to.

Oracle Database@ cloud service offerings also enjoy the same high levels of scalability and availability as Oracle Database on OCI, as well as feature and price parity. Exadata optimizations for transaction processing, analytics, AI, and mixed workloads help ensure there are significant performance benefits.

That brings me to latency, the time it takes for data to traverse the network. Single-digit millisecond performance is sufficient for many use cases, but not all. Colocating Oracle Exadata systems within cloud providers’ data centers makes it possible to reduce latency to microseconds.

“This streamlines interactions between the Oracle Database and the Azure middle or application tier, cutting latency in some cases from milliseconds down to microseconds,” said Oracle SVP, Product Management Jenny Tsai-Smith, in explaining the advantages of Oracle Database@Azure. The same can be said for Oracle Database@AWS and Oracle Database@Google Cloud.

Data management upside

The ability to run Oracle Database in AWS, Azure, and Google Cloud comes with other benefits that help companies make better use of their data.

Here’s an example: Extract, transform, and load, or ETL, is a long-established norm in data curation and warehousing. But this kind of data normalization and cleansing is typically a time-consuming process. More recently, “zero ETL” technology uses APIs and other integrations to automatically position data where it’s needed for analysis without the front-end work—or better yet not move it all and rather serve OLTP, analytical, and AI needs from the same converged Oracle Database that is capable of running any workload type and supporting any data model.

The Oracle-AWS partnership enables zero ETL integration between Oracle Database services and AWS analytics services. Thus, customers can analyze data across Oracle Database services and applications that are running in AWS without having to build and maintain complex data pipelines. It’s faster and more efficient as customers don’t have to manage multiple copies of data.

Or consider backup and restore operations. Native integration with AWS means you can back up, restore, and initiate other disaster recovery actions to Amazon S3 storage without extra effort.

Tools of choice for AI development

For tech teams looking to fast-track AI projects, I saved the best for last. In earlier columns, I’ve written about the new AI capabilities built into Oracle Database 23ai—machine learning algorithms, natural language queries, and vector embedding and unstructured data search via AI Vector Search.

Not only do customers now gain access to these and other Oracle AI innovations in AWS, Azure, and Google Cloud, but it’s a two-way street: They can tap into the enterprise data in their Oracle Databases using popular AI development platforms, like Google Cloud Vertex AI, Microsoft Azure AI, and Amazon Bedrock.

At Oracle CloudWorld in Las Vegas, I attended a session where executives from Accenture, Google Cloud, and Oracle discussed how this works.

“I’m excited by the GenAI engagements we have going on,” said Andi Gutmans, Google Cloud VP and general manager of databases. “All of these customers also have Oracle systems. These transformative GenAI use cases become so much easier. It’s all about the data.”

The AI-enabling integrations don’t end there. Developers can tap into Oracle Databases using Google Cloud services, such as Gemini foundation models. In Azure, they can mingle Azure AI services, like OpenAI large language models, with Oracle enterprise data. Or in AWS, they can seamlessly connect enterprise data in their Oracle Database to AWS’s AI and ML services, including Amazon Bedrock.

Building on today’s multicloud foundation

I’ve only scratched the surface of the possibilities created by these still-expanding partnerships. Other new capabilities include single sign-on for improved cross-cloud security, data architectures that span availability zones, and streamlined database migrations. You can even do cloud hopping, say by consolidating Amazon RDS on Oracle Database@Azure.

And Oracle HeatWave capabilities, including generative AI, lakehouse, and machine learning, are available on AWS and Azure. In fact, a customer in the tech industry at Oracle CloudWorld noted that they went into production on HeatWave GenAI in less than 30 days. So the story expands to multiple databases on multiple clouds.

And this is just the beginning. The new spirit of customer-focused cooperation among the world’s four leading hyperscalers lays the groundwork for continued multicloud innovations. It’s an encouraging sign of things to come for developer teams building in the cloud and for business leaders looking to seize data-driven opportunities—in the words of Larry Ellison—anywhere and everywhere.

John Foley is editor of the Cloud Database Report and a vice president with Method Communications.

View more Oracle Connect articles