Oracle unveils final cog in cloud strategy

The final piece of Oracle’s cloud strategy was unveiled last week with its Roving Edge Infrastructure, a ruggedised device that packs computing, networking and storage capabilities into a single box to run workloads in remote and edge locations.

The $160-a-day device, which lets enterprises run cloud applications with low-latency requirements at the edge as an extension of Oracle Cloud Infrastructure (OCI), is already seeing interest from some enterprises in the Asia-Pacific region.

“We’ve got mining companies in Australia with offshore and remote operations that are looking at this,” said Chris Chelliah, senior vice-president for customer strategy, business development and insight at Oracle in Japan and Asia-Pacific.

“We’ve also got a couple of manufacturers in Korea and ASEAN that are looking to do simulations and capture instrumentation data on the shop floor before shipping it to their mothership where they’ll run machine learning algorithms against it,” he said.

Other industries that might benefit from the Roving Edge Infrastructure, said Chelliah, include defence and agriculture where the device can be used to process data locally before tapping public cloud services for further analysis.

However, as an infrastructure centric device, Roving Edge Infrastructure does not come with cloud platform services.

So, to perform machine learning inference at the edge, for example, enterprises can deploy their own inferencing engines or use an algorithm from independent software vendors in the Oracle Cloud Marketplace.

Oracle’s Roving Edge Infrastructure should be seen in the context of its broader strategy to go after workloads that had been stubbornly difficult to move to public cloud services, whether they are workloads that deal with sensitive data or those with security requirements.

Chelliah said although Oracle has acknowledged many times that it was late to the cloud game, the company is going after tougher problems.

“We’re solving the problems for stuff that still hasn't moved to public cloud because of requirements such as isolation and the ability to pick stuff up,” he added, referring to the ability to run and manage workloads on the cloud as they were on-premise.

Then, there are also VMware workloads that Oracle and major public cloud providers have been eager to host through various tie-ups with VMware itself.

Chelliah said with its VMware offering, Oracle has been able to “pick up customers’ entire VMware states and run that on our cloud” to support disaster recovery, which has been one of the top three use cases in ASEAN, Korea and Taiwan.

“Now, that’s very different to how Amazon, Microsoft and Google do it – they offer it as a managed service, which is a completely different environment and configuration, whereas we’re making it sort of native and it’s a natural extension to the old process.”

Chelliah said Oracle’s Dedicated Region Cloud@Customer has also seen successes with organisations that wish to run an entire public cloud region on-premise.

In Australia, Australian Data Centres (ADC) recently said that it would deploy Oracle’s Dedicated Region Cloud@Customer to host cloud services for the federal government and meet data sovereignty and security requirements.

Through the Oracle managed service, ADC will be able access all of Oracle’s second-generation cloud services, including software-as-a-service applications, bare metal compute, autonomous database, as well as container-based services while maintaining control and governance of their systems and services.