Oracle's cloudy Spice Girls moment - tell us what you want, what you really, really want

Martin Banks Profile picture for user mbanks February 15, 2017
Oracle can talk tech-and-dirty if you want to, but now it seems it would rather take a more holistic, business goals approach to the whole subject of the cloud. 

Whatever cloud word with a blue sky © gustavofrazao -
Oracle may have come to the cloud qua cloud a bit later than many, Larry Ellison having been one of its more vocal detractors early on - water vapor, anyone?

But since its Pauline conversion in 2012 it has made some significant strides, to the point where it contends to offer a plan that maps onto its market demographic better than many a CIO might have thought possible.

It is also a plan that marks a counterpoint to the idea being promoted by Amazon’s AWS, and to a lesser extent the likes of Google and Microsoft Azure. That is the notion of size being the key differentiation in the delivery of cloud services. Size, the shear physical capacity of the resources available to users, is promoted as the key marker in defining the right choice of cloud service provider.

Now, up to a point, that is true – especially if the potential customer is a major corporate, the type of business that might well be basing much of its mission critical processes on Oracle databases and applications. Size equals resource headroom, which in turn equals reliability of service provision. It also means competitive strength, especially in the area of price.

But price is not everything. Increasingly, what is of more value is everything that goes to make up that nebulous concept – the service. In particular, anything that reduces the workload on a business of getting that service delivering valuable business results has the potential can cut set-up and engineering costs, and hasten the all-important time to cash.

In that context – a platform geared at least to the needs of a market sector if not, as in Oracle’s case, the needs of users looking for comprehensive levels of support for their plans and attempts to move existing Oracle-based business systems into the cloud – price becomes a far less relevant factor.

More important is the need to transition flexibly, reliably and with as little engineering as possible to a cloud environment that provides value in as short a time as possible.

When lots of  'smalls' can be better than one 'big'

A useful conversation with Steve Daheb, Senior Vice President for Oracle Cloud, at the recent Oracle Modern Business Experience event in London gave me some insight into the direction Oracle is travelling to deliver such environments.

One of the most obvious steps of late is the significant increase in Oracle data centers that will be available to its customers. Announced in mid-January, the additional resources now bring the total number of geographic areas covered to 29. This addresses one of the key underlying questions that follows on from the model of having a few, extremely large datacentre resource centres available to serve the world at a good price.

There is a school of thought here which suggests it is better to have a larger number of smaller, more-focused datacenters scattered around the world, and this is a school to which Oracle subscribes. This way, it should become possible to combine commonality of technology, code and service support delivered together with better performance and reduced service latency. It also means that two other service factors can be met - data sovereignty and localization, as Daheb indicated:

There are obviously some sets of services that can be homogenous on a global basis. But I think there are other drivers for why you maybe want to have to datacentres closer to the customer. Sovereignty is coming up a lot, as is latency, but either way you're going to be looking at some of the newer IaaS offerings that we're rolling out around pure metal and software-defined virtual.

He even pondered the question as to whether some of that support effort even classifies as being part of the platform, though it is part of the service. There are multiple layers of differentiation to accommodate, ranging from company offices operating in different countries and regions, or the way that different industries requires applications to be optimized in different ways.

It's not really core to the data centre, but there's other ways to differentiate for particular markets and segments.

Cloud is holisticism, not a tech

Daheb feels that Oracle looks at the cloud in a far more holistic way than many of its competitors. Yes, it operates as an IaaS, PaaS and SaaS provider, with IaaS providing the core infrastructure and management for scalability, PaaS providing the delivery, configuration and optimization of the database, the integration, the analytics and the security, and a SaaS layer based around the Oracle applications suite already widely used on-premise by corporations. It also has the networking component as well as server hardware technology on hand. It can therefore connect in discussion with customers at any level desired:

We could start with a SaaS app, but you want to connect that back to what you have on prem, and that's where the PaaS layer comes in from an integration perspective, and then maybe with respect to data capture or scale or other things that can be leveraging IaaS.

He said the company looks at the stack from an applications perspective, from platform down to infrastructure. At the same time it also looks at it from a hybrid perspective as well because there are a lot of customers looking at their world from the perspective of new workloads that need to be accommodated. So flexibility of operations is important, with users able to choose how they want to approach workloads. For example, they can develop in the cloud and run production on-premise or run it the cloud and connect back to on-premise.

A classic workload is to run in the cloud when the workloads are light but when it comes to running core financials using all the 10,000 customisations developed over many years, run it on premise where the optimisation is already pre-configured:

So when we look at customers converting to cloud, there's so many different entry points. It could be SaaS, PaaS or Just Compute, Cloud native app dev, or may contain things from an IaaS perspective. This is where we can re-frame or define what cloud or cloud services means from a broader perspective and then ask how does that relate to different customers.

And what of NetSuite - Oracle's recent cloud ERP acquisition - in all this? One interesting factor in amongst this holistic approach is that it does not – at least for the publicly acknowledgeable future – stretch to cover NetSuite. That company’s former-CEO, now EVP of Cloud within Oracle, Zach Nelson, was a keynote speaker at the London conference, and stated that the company had over 20,000 customers. Daheb stated that Oracle also had some 20,000 SaaS customers.

What is more, he suggested that a good proportion of those were net-new customers to Oracle that had moved to sign up for the SaaS-delivered services and were not previous on-premise users. One of the targets for Oracle’s SaaS offering is now the SMB marketplace, which is where NetSuite largely plays best. There are other similarities as well.

The database being targeted at the SMB sector by the company is Exadata, and Daheb stated that the SaaS delivered version for that market, Exadata Express, starts at around $179 a month, which he claimed is less expensive than the equivalent offering from AWS. The Exadata database is however a development that has its routes in the Oracle database technology and therefore useful compatibility not available with the AWS offering.

NetSuite has its own database, but according to Daheb this too is based on Oracle technology, which gives the company the long term possibility for extending the available holisticity across the product range that users can employ:

Exadata is the database, and NetSuite is at the application layer. Well when you buy the application layer typically that is in the cloud. You have the database and you have associated storage and compute built in. you also have people that are doing their own custom apps, different applications where you can use an Exadata or the cloud, you can use Exadata express and there's that stack that's tied into it. So it's all congruent and complementary.

My take

This does indicate that Oracle is making the sensible move of avoiding much of the technology bragging common amongst cloud service providers, and instead pitching itself at the more business-related question: what do Oracle users actually want to do? By going for that business-solution-first model it does seem as though the company is gaining a useful side benefit – SMB first-timers - businesses new to Oracle that might just grow into large scale corporate users.

A grey colored placeholder image