OpenWorld Throughout this year's Oracle OpenWorld conference, the company's message has been fairly simple – the database firm is moving to cloud but can offer buyers either on-prem kit, full public cloud, or anything in-between.
It's a tempting pitch for those who are somewhat nervous about diving into the cloud with their systems, and for those existing customers who aren't particularly happy at the prospect of trusting Amazon, Azure, or any other vendor with their mission critical systems. But will it work?
For Oracle the pitch is ideal. The firm was late acknowledging the cloud as a business driver (indeed Ellison was openly contemptuous of the possibility not so many years ago), but has devoted billions in spending to bring its apps and infrastructure onto a cloudy footing.
But by offering itself as a cross-site vendor, Oracle could win on two fronts. Firstly it keeps its on-premises business, which is very profitable. At the same time it can woo the cloud-friendly market with either a public cloud, or with a hybrid approach it calls Oracle Cloud at Customer.
That last option bears some examination. The company says customers get the best of both worlds – the scalability of cloud and the metered pricing that allows quick scale-up and -down of capacity as needed – but keeps the entire thing behind a customer's firewall and reasonably safe.
There are some similarities here with Microsoft's approach to operating systems and applications. Redmond has been selling itself as one of the few companies with experience in both public and private cloud, as well as hybrid systems taking elements of both, and has enjoyed some success with the tactic.
There's no reason to think that Oracle can't do something similar with its customers in the future. Certainly Oracle wants to see it going that way, and has been making some bold (and occasionally misguided) predictions as to how it's going to take the cloud market by storm.
Amazon doesn't need to worry for the moment
So should other cloudy vendors be quaking in their stylish yet affordable boots? Not yet, and maybe not ever.
Larry Ellison and his cohorts have been aiming their guns at Amazon, saying that Oracle wants to challenge Bezos' mob directly and outpace AWS in market share and revenue. One would imagine the reaction from Amazon is, "That's cute, now run along and play with your boats."
On the face of it there's no contest. AWS sales and revenues dwarf those of Oracle, measured in the billions, not Larry's hundreds of millions. Amazon pretty much created the popular cloud market, drew millions of customers in, and shows no sign of losing them.
Oracle, by contrast, got cloud wrong. Ellison saw the potential out there – but thought it would be a return to client/server and so wasted four years with toys like the network computer. When that didn't work, he eschewed the cloud, only to acknowledge his mistake and then throw lots of money at the problem.
Based on what we've seen this week, the firm has done an excellent job in translating its code base for cloud use, and can certainly offer a reliable service for those wanting a hybrid approach of on-prem kit and public cloud scalability.
Oracle has even taken aim at its rivals on pricing, undercutting AWS for some use cases. But that's a risky strategy – far too many Oracle customers in the past have found themselves facing big bills from the company and may be unwilling to trust the firm again.
In the meantime, Microsoft is cutting into Oracle's turf with DocumentDB, Amazon is sitting pretty, and there are also open source database alternatives that are building in popularity. Oracle is well positioned to keep bringing in big revenues with cloud, and maintain its on-prem business, but it's no danger to the big boys in the short or medium term.
"Amazon has nothing to worry about yet. Blockbuster and Nokia once also had nothing to worry about," blogged Marcel van den Berg, a Microsoft Azure and VMware vSphere expert.
"Many vendors tried to compete with Amazon and Azure. For example VMware did an attempt to become a global infrastructure-as-a-service (IaaS) vendor with vCloud Air. VMware decided to stop further development of the IaaS platform and partner with IBM instead. HP tried public cloud with Helion. They stopped as well. Verizon terminated the Public Cloud offering for IaaS. While the exact reason is not made public, it is safe to assume the service could not be made profitable in an acceptable timeframe.
"Oracle's [second generation IaaS] will have a hard time to become a leader in IaaS for general purpose workloads. It will be next to impossible to catch up with Amazon and Azure on technology and marketshare. More likely Oracle will become a niche player for high performance compute." ®