Oracle and OpenStack: A Tale of Two Completely Opposite Strategies

Oracle and OpenStack: A Tale of Two Completely Opposite Strategies

Summary: If you wanted to see the full spectrum of cloud choices that are coming to market today you only have to look at these two efforts as they are starting to evolve.

SHARE:

If you wanted to see the full spectrum of cloud choices that are coming to market today you only have to look at these two efforts as they are starting to evolve. They represent the extremes. And ironically both held analyst events this week.

OpenStack is clearly an effort by a vendor (Rackspace) to launch a community to help advance technology and drive innovation around a framework that multiple vendors can use to bring myriad cloud services to market and deliver differentiated values. Whereas Oracle, who gave analysts a brief look inside itspublic cloud efforts this week, is taking a completely closed and self-built approach that looks to fulfill all cloud values from top to bottom.

OpenStack arrived on the scene roughly two years ago as a very incomplete set of cloud capabilities – Swift, an object store, lifted straight from Rackspace Cloud, and Nova, a rudimentary, but scalable compute service developed by NASA. Since it appeared on the scene it has gained significant momentum in the user (early adopter) and vendor communities and we’ve seen rapid iteration and hardening of the base framework. Earlier this month its 5th release,Essex, was made available for download. And since its third release (Cactus) it has seen multiple implementations at hosters (InternapSoftLayer,DreamhostSina.com and most recently HP Cloud), scientific institutions (San Diego Supercomputer CenterNeCTAR and Argonne National Laboratory) and enterprises (Mercado Libre, Sony Playstation and others). It is by no means a finished product in its current state but does provide the base capabilities upon which a finished service can be built. That’s certainly what Piston Cloud and MorphLabs have already done. Red HatHP,NebulaDell and several other firms have announced plans to do the same in the future (most likely using the 6th release, Folsom, which is expected to be available in September 2012). The focus of this effort has been around growing the community and driving collaborative iteration of the technology.Hundreds of vendors have contributed code, helped launch additional projects and all of them see OpenStack as a mechanism for go to market advantage. Members also see the overall community ecosystem as something that can fuel the success of all participants. The effort is solely focused on delivering foundational cloud capabilities – Infrastructure as a Service (IaaS) -- but vendors with designs on delivering higher values (PaaS, SaaS and other cloud capabilities) are jumping on the OpenStack bandwagon as it provides the opportunity for a much larger market opportunity than being dependent on one or a few cloud solutions or providers’ proprietary stacks.

Way at the other end of the spectrum you have Oracle’s cloud efforts which center on leveraging all its vast intellectual property to build a very Oracle-specific set of capabilities. Much of what Oracle shared with analysts is confidential – protected by a non-disclosure agreement until it will be publically unveiled later this year – but what can be shared shows an effort to tell the world, “we have all the technology necessary to provide a full suite of cloud services and we intend to deliver them to our customers on our own.” There are some similarities to Microsoft’s early cloud strategy that was called “Software + Services” in that Oracle clearly wants to provide customers the choice of deploying its applications and other IP either on-premise or in the cloud and are striving through the use of Fusion middleware to make the transition between these deployment types as seamless as possible. Oracle is also eschewing IaaS and offer only PaaS and a series of SaaS solutions – some of which are already in market.

However, its approach is actually most similar to Salesforce.com’s strategy than Microsoft, as Oracle is an applications company first and sees its SaaS versions of its Fusion Apps as the main point of entry. Its PaaS appears to be a place to deploy customizations and add-ons to its SaaS services, much the wayForce.com is used by Salesforce CRM customers today. And low and behold Oracle has database and social media cloud services planned as well.

Let’s looks at its cloud services layers in a bit more detail:

It’s PaaS layer is based on the WebLogic Application Platform and will provide Java EE support as you would expect. However it won’t be a Java-only PaaS as, according to Oracle Executive Vice President, Thomas Kurian, it will support PHP, Python, Ruby and Rails applications also, via a web service. There will also be a web sites services (a la Force.com Sites) and content management and integration services. The whole platform will run atop ExaLogic hardwareso it should take advantage of the solid-state disc and InfiniBand network backbones within these appliances to deliver strong performance. I’m sure there will be a “We’re 15X faster than Force.com” ad campaign soon.

On the applications side, Oracle could have simply chosen to expose all its Fusion Applications and its acquired SaaS applications, like Taleo andRightNow, and let customers put together their own workflows, but Oracle decided it would have better success doing this integration itself. So the company said it plans to launch with complete business processes for enterprise resource planning (ERP), Human Capital & Talent Management (Taleo integrated with existing Oracle Fusion apps), B2B and B2C Customer Experience services and Supply Chain Management (SCM). Each of these services will weave together a variety of Oracle applications into a consistent workflow and business process, which is clearly a more ambitious and likely more user-friendly approach than a Chinese menu of Oracle SaaS apps. The ERP SaaS service, for example weaves together Oracle Hyperion and Primavera, Fusion Procurement, Financial Intelligence and Accounting, and other Oracle applications.

Oracle didn’t disclose launch timeframes we can share with you but if you are interested you can request a log-in to its early access program today at cloud.oracle.com. Certainly we’ll hear much more about this at Oracle Open World in the fall.

What stands in stark contrast to OpenStack is the completeness and single-company approach Oracle is taking to delivering cloud services. While interfaces to the cloud will be open and multiple developer frameworks will be supported this is clearly an all-Oracle solution. Not a single ISV partner was mentioned during any of the discussions about this cloud effort (unless you count all the ISVs they have acquired).

This aligns very clearly with Oracle’s strategy over the past two years since it acquired Sun Microsystems – deliver a complete stack solution designed and optimized to work together. Oracle is essentially saying, “It’s integratable with other vendor products – but why would you?”

You can’t get any more dichotomy between cloud strategies than these.

Topics: Data Centers, Apps, Cloud, Emerging Tech, Oracle, Software Development

James Staten

About James Staten

James Staten is a Vice President and Principal Analyst at Forrester Research, serving Infrastructure and Operations professionals.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

1 comment
Log in or register to join the discussion
  • Lock yourself into a smelly prison

    Wow! Given the commercial practices of Oracle, who would want to lock themselves into an inexperienced cloud company? I will agree that their database is good, but that has nothing to do with cloud capabilities. Their customers are either going to be lemmings or moths circulating around a candle.
    HardTruth