It’s a multi-cloud world and one with a cloud infrastructure services market that is dominated by three large players. If you’re one of those cloud services providers that is not among the big three, how do you expand the reach of your own cloud offerings?
Well, on strategy is that if you can’t beat ‘em, co-opt them for your own designs.
That is what VMware, which was not rich enough to build its own clouds, tried to do with its strong partnership with AWS. VMware ported its vSphere server virtualization hypervisor and its related software-defined datacenter stack to AWS servers, storage, and networks back in 2016 and sold it as a VMware Cloud on AWS, a separate and distinct cloud operated by the world’s largest cloud builder. (This was the brainchild of Pat Gelsinger, who was chief executive officer of VMware at the time, after he realized that VMware did not have anywhere near the cash necessary to build a rival cloud.)
Oracle is taking a slightly different – and we think what we ultimately be more successful – tack as it partners with its cloud rivals to extend its reach. Oracle will not only offer its database and other services on its own cloud infrastructure, but is also partnering with its competitors, who have much larger fleets of infrastructure around the world, to get those same Oracle services on that cloud infrastructure.
Over the past year or so, Oracle has been enabling enterprises to run Oracle database services on Oracle Cloud Infrastructure (OCI) inside Azure datacenters, a way to give organizations more flexibility in the cloud while giving them more avenues to access OCI database services for their workloads.
In June, Oracle announced a similar partnership with Google to allow companies to combine both OCI and Google Cloud technologies. The two companies unveiled Oracle Database@Google Cloud, making it easier for users to accelerate the migration of their workloads to the cloud and to take advantage of not only Google Cloud’s infrastructure but also its tools and AI services, such as Vertex AI and its Gemini generative AI models.
As Oracle CloudWorld 2024 opened this week, company executives not only said that Oracle Database@Google Cloud is generally available, but that a similar collaboration with Amazon Web Services also is available to enterprises.
With Google, the partnership is allowing organizations to run Oracle services like Exadata Database, Autonomous Database, and Database Zero Data Loss Autonomous Recovery services – which creates near real-time backups, which is important during a cyberattack – on OCI in the Google Cloud datacenters. As with almost everything these days, the services will be replete with AI capabilities. The Exadata and Autonomous Database services will be running Oracle Database 23ai, so all those capabilities will be available to users, according to Leo Leung, group vice president for Oracle Tech and OCI.
In addition, users can import Oracle Linux images using Google Cloud’s virtual disk image import process and, within the next year will be able to leverage ready-to-use images to streamline Oracle Linux image provisioning in Google Compute Engine.
“By doing that, we’re physically co-locating OCI next to Google’s infrastructure and services, which eliminates any kind of latency challenges,” Leung told The Next Platform before Oracle CloudWorld opened in Las Vegas. “Then we’re integrating it into the Google control plane so that customers can go into their Google Cloud Console, they can use the APIs and these services then feel a lot like native Google services so they can work side by side.
He added that “the goal is to make these services native right next to Google services, where you can start to combine them with, for example, the Google Kubernetes Engine to run your application, or Vertex AI in order to take your model in Vertex AI and be able to use it against data inside of the database in Google Cloud, use it against the vector search. Really combine these services for a number of migration as well as innovation types of use cases.”
With AWS, Oracle is launching Oracle Database@AWS, which allows customers to access Oracle Autonomous Database on dedicated infrastructure and Oracle Exadata Database Service within AWS. Oracle Database@AWS will connect enterprise data in Oracle Database and workloads running on Amazon EC2, AWS Analytics services, and the cloud provider’s AI and machine learning services like Amazon’s Bedrock fully managed service.
It also delivers a low-latency network connection between Oracle databases and applications running on AWS.
Multi-cloud is a fact of life in the business world, with 89 percent of organizations surveyed by Flexera in the most recent of its annual State of the Cloud report using it. And according to numbers from Synergy Research Group last month, AWS continues to be the largest cloud infrastructure services vendor, with 32 percent of the market, followed by Azure with 23 percent, and Google Cloud at 12 percent.
That said, Oracle joined Huawei, Snowflake, and MongoDB as fast-rising followers, with Oracle surpassing IBM for the first time. In its latest quarterly earnings released September 9, the company saw its cloud revenue jump 21 percent year-over-year, to $5.6 billion.
In a statement, Oracle chief executive officer Safra Catz noted that cloud services is Oracle’s largest business and touted the agreement with AWS.
“It’s about multi-cloud. It’s about choice,” Leung said. “Customers for a long time, on-prem, had choice. They believed in best-of-breed. They used multiple vendors together to form their solutions. The cloud broke that for a while, where it was all walled gardens and you had to consume only from a single cloud and that was it. This really breaks it down so that customers have the choice they want. It will dramatically increase cloud adoption overall because a lot of these types of workloads are still stuck on-prem.”
Microsoft, Amazon, and Google have no shortage of tech partners for their cloud operations, so Oracle isn’t unique in that sense. However, Leung argued that how the partnerships with Oracle work is different. Typically, enterprises can get access to, say, Oracle Database, but it would have to run on the cloud provider’s compute.
“You would have to accept all the compromises that come with that,” he said. “The same thing goes with a lot of the third parties, and you would never be really be able to get the full capabilities of those third parties. Or you would run a cloud across multiple clouds and then have to figure out the management across those two clouds and the security across the commercials and billing [for] two or three or four clouds. Almost every customer I talked to has at least two, if not three or four cloud vendors. It’s just very, very complex, either complex or somehow compromised.”
The way Oracle is addressing it with Microsoft, Google, and AWS – being able to run OCI within the cloud environments – “really tackles that head-on. It’s the full capabilities. There are no performance compromises. There’s no commercial compromise. There’s no operational compromise, so it’s a very different approach.”
Leung said Oracle is seeing organizations embracing what it’s doing. The company noted that MSCI, which offers support tools and services to global investors, and New Zealand-based multinational Fronterra are both adopting Oracle Database@Azure. It is available now in Google Cloud datacenters in four regions in the United States, Germany, and the United Kingdom, with plans to expand it to other sites in around the world, including the Middle East, Latin America, and Africa. Oracle has 162 cloud datacenters in operation and under construction around the world.