Why is Oracle’s cloud strategy such a confusing mess?

Oracle is a different company to the one many have grown up with – it is no longer a database company that just happens to have a few applications tagged on.

The acquisitions of the last few years have made it into a major applications force, with Siebel, Peoplesoft, JD Edwards being the major purchases, along with the likes of Retek, BEA Systems, Primavera, Hyperion and Stellent adding capabilities to fill holes in the portfolio. In fact, Oracle has made over 65 acquisitions in the last 10 years, changing the way that it has had to operate and what it can offer to the markets.

The biggest change that surprised many, though, was the acquisition of Sun Microsystems in January 2010. A supposed software-only company (OK – it had tried Raw Iron in the late 1990s, but had found that it was not well suited to such a market at the time) suddenly decided that it wanted to be a hardware and operating system company – but why?

The likes of IBM and HP were making big strides in trying to become the big stack vendors – selling companies the platform from which they would run any applications, from the hardware and operating system to the middleware and other components needed in an enterprise environment.

This threatened Oracle – IBM has DB2 as a major competitor to its database (currently Oracle 11g), and also had bought Cognos and SPSS for business intelligence that would compete with Oracle’s own Hyperion acquisition. IBM also has WebSphere, the biggest competitor to Oracle’s BEA WebLogic middleware/application server platform.

HP, while having been a little less successful with its own acquisitions in areas such as middleware (Bluestone came and went without many people noticing), still could muster a credible platform with its own hardware, storage, networking and operating systems, along with large parts of the supporting software required. Even Dell and Fujitsu were beginning to make noises at making more of a play to be the de facto supplier of a complete IT platform to large enterprises – and Oracle had to act to defend its own position.

Cloud has also been a major force that has helped Oracle define its future. Sun had made its name in the dot com boom (“We put the dot in dot com”), and Oracle perceives that Sun’s heritage positions it well for the cloud boom.

Sun, having made its own acquisitions in the years before Oracle came to the rescue, not only had server hardware and a well-proven and well-loved operating system, but also had its own networking hardware and brought in storage systems through its acquisition of StorageTek.

Through the Oracle acquisition, Oracle now was in the position to bring together everything from the general compute, network and storage hardware through the operating system to application server and middleware and on to the database and application levels, bolstering all of this with systems management, business intelligence and other systems on the one platform. Since the Sun acquisition, Oracle has further boosted its storage offerings through the acquisition of Pillar Data Systems.

Surely, a recipe for instant success? Well, not necessarily. It has to be acknowledged that Sun retains a loyal, but somewhat shrunken user base. This base lies in just a few verticals, such as finance, pharmaceuticals and telecoms. Oracle brings in other verticals, such as retail and public sector, but just how can Oracle play its stack as part of the general cloud market?

It is unlikely that a general commercial organisation looking to move to the cloud will have Oracle at the top of its list. Although Oracle will also provide support for Linux on the stack, the use of Solaris as its own operating system minimises the appeal for a mixed workload environment. Here, IBM and HP have better solutions that cover Wintel and Lintel, as well as more targeted workloads on their own RISC Unix platforms. Dell provides its own scale out Wintel/Lintel systems for general computing.

But it may not be all bad news for Oracle – a homogeneous, high performant stack is something that many cloud providers need, and although the likes of Amazon and VirtuStream provide an IaaS/PaaS offering here, the independents are still facing a dynamic market, with no 500 pound gorilla having taken the market by storm.

Microsoft is playing Azure hard, Nimbula is offering its cloud operating system as an Amazon EC2 alternative but both of these are not stack offerings, but just the software layer. New entrants into the cloud market will be looking for a far more complete offering – and Oracle will be competing directly with IBM, and to a lesser extent HP and Dell here.

Provided that Oracle can play the its stack intelligently to a targeted prospective market, and can make sure that a multi-workload environment is catered for through intelligent provisioning and routing of workloads to either a Solaris or Linux platform, it will appeal to cloud providers looking to building a predominantly open source platform. Oracle can sell the platform as a complete entity, with its database, business intelligence, middleware and so on all as part of that system.

Within the private data centre, Oracle would do better to take a Cisco UCS approach – creating a highly modularised compute “block” that can be run alongside other systems, so enabling the Oracle Stack block to be highly tuned for specific workloads. Already, Oracle has launched the Exalogic Elastic Cloud appliance – but if anything, this may actually confuse the market as it is only part of an overall cloud approach, not really the “cloud in a box” that many seem to believe that it is.

Overall, Oracle’s cloud strategy is confused. It is making a lot more noise around its own cloud services, with its own applications being made available from Oracle owned and run data centre facilities, but it also needs to define exactly where it expects to play in the commercial cloud markets through being a provider of a total stack. Without crisper messaging and a more defined offer, Oracle runs the risk of being trampled under the feet of others making stronger plays for this incredibly important market.

SHARETweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Pin on PinterestDigg thisShare on RedditShare on TumblrShare on StumbleUponEmail this to someone

Clive Longbottom is founder of Quocirca and is a highly respected and globally recognised industry analyst, covering a range of business and technology areas. Clive’s primary coverage area is business process facilitation. Clive has been an ITC industry analyst for over 15 years. Clive has worked with a range of large and small analyst companies, including META Group (now Gartner) as VP Europe. Clive has a B.Sc. (Hons) in Chemical Engineering from the University of Aston in the UK.