IT Management Blog: my thoughts about putting the "i" in IT

Why IT Managers are getting tired of "the Cloud"

I noticed recently that CIO's and IT professionals will roll their eyes when you mention the word "cloud" as if they had enough of the term and the hype around it.

Though most of us all acknowledge that these new technologies and services provide great opportunities, in general most feel that the term is overused and that in many cases reality is not as simple as vendors seem to make it be.

What the CIO's and IT professionals see is that the world goes around in circles and that they need to explain the hidden complexities of the reality of cloud computing. And this is not easy.

How foggy is the cloud?
The term cloud is used almost for everything and there are many definitions going around. First of all we see it being referred to as the Software as a Service solution such as Google Apps or Salesforce. But on the other extreme I heard it to being referred to as your internal virtualisation of your own servers in the context as private cloud.

Gartner defines "Cloud Computing" as "A style of computing where scalable and elastic IT-enabled capabilities are delivered as a service to customers using Internet technologies."

And it defines "Private Cloud" as "a form of cloud computing where service access is limited or the customer has some control/ownership of the service implementation."

How generic are those definitions? "Using Internet technologies"?  Yes, I can use Intenet technologies on my standalone computer network at home and have it disconnected from anything else in the world. Are we going to call that cloud computing or private cloud?

For example I have been explained that one organisation wanted to have a cloud solution but wanted to keep the data within its own walls. The vendor therefore placed it's own servers with the application in the datacenter of it's client and managed the application remotely.  Isn't that just a managed service? What's so cloudy about this?

Another definiton can be found here: http://en.wikipedia.org/wiki/Cloud_computing

Personally, my opinion is that many of the cloud options are simply variations to the traditional outsourcing that probably already existed since the first computer was created. I remember when I just started at Uni in the 80's that the mainframes we did our programming exercises on were owned and managed by the Academic Hospital across the road. SaaS was probably also introduced immediately with the inception of the Internet.

What I call a pure cloud service is when you access it over the public Internet, that everything you provide is standardised and, except for configuration, it is all exactly the same for other customers. As soon as you start customising the standard offering, your service moves into a different category.

The question remains than what it is that you buy in the cloud: data center, servers, operating system, application, managed support services, development services?

So what is the issue with the cloud?
So far so good. We introduced a new term for something that existed already for a while. But why do IT people get this tired look in their eyes when you mention the cloud?

One of the main challenges IT professionals within end-user organisations face, is the expectations of the business and executives have of the cloud. They see and hear the great stories of the vendors and see how easy it is to use Google Apps or Gmail. They hear about success stories of organisations who put all their email and and document management in the cloud. So why can't we do this? Then they have the experience with some offerings for example with the HR or Payroll systems where the application was easily implemented, hosted and supported by the vendor. So why can't we put the ERP system in the cloud? Or not all our infrastructure?

It is not about the legal issues and in which jurisdiction the data resides. Though these are issues IT managers will need to chase up, in the end it is considered to be a legal issue that the business needs to resolve.

Security is of course another concern. But let's assume that those issues can be overcome.

Payroll is for many organisations a standard process. The system, except for some configuration, does not require customisation and in fact the whole business process for Payroll can often easily be outsourced. But for the other business systems it is not that easy. Yes, Oracle would like to take ownership of your ERP system and they can provide services for that. But your ERP hardly ever stands alone and you will always have customisations. Handing over your ERP system to a vendor is not per se about where the system resides, but it is all about the managed services.

Due to the size of many ERP installations, a shared infrastructure does usually not apply and if you would do this, it would probably require a signficant modification to your system to accomodate for that. What you are left with is a shared data center.

Due to the customisations, a shared software installation of the ERP system would not be feasible and much of the real cloud benefits cannot be achieved.

Due to the size, complexity, customisations and integration with other systems, the vendor would need to setup almost a dedicated team to provide the support and development services for you.

What you are left with is still a dedicated system, hosted in another location and with a dedicated team to provide the services. In the end you need to evaluate all options against its merits which is not much different as we've been doing over the many years.

Another key issue is, and that is more of a concern to me, is that if you put a of variety systems in the cloud, you suddenly have a large series of vendors to manage. And you might expect that these systems one way or the other integrate. But what do you do when vendor A wants to upgrade their system and the solution managed by vendor B then also needs to change? It means all the iniatives of the various vendors need to be synchronised. You'll need quite a team to manage all this regardless of all the contractual issues to be addressed.

And what will you do with a custom built solution that tightly integrates with your ERP system? Say, you put your ERP in the cloud and it will be located somewhere at the other side of the world. Your bespoke solution requires each time to request data with twice the latency across the globe which will signficantly impact the performance of the bespoke system. Besides, if this access needs to happen over the public Internet, you will run into some reliability issues.

Besides the many problems technically, I also see that it over time it might financially not turn out to be optimal.

So why were they rolling their eyes again?
The problem that CIO's and IT managers face is trying to explain the complexities to the business without being seen as too protective and too resistant to change. They expect the world to go around in circles again. We've seen it with the client-server technologies. First everything was centralised, then decentralised and now centralised again. We probably start working with many different cloud providers. This then becomes inefficient and we then need to bring that together again.

Is it all that bad?
A good option of the cloud offerings are is that many business managers can put the serves in place without support of IT. This can really empower the business and make businesses much more agile. The flip side of this, is that organisations will have many contracts with vendors and that the left hand will not know what the right hand is doing.  It might lead to a messy situation and subsequently over time a more centrally managed approach will be required.

But I do see a lot of good coming from the cloud services that emerge everywhere. First of all small and medium business have now access to solutions they otherwise would not be able to afford. I expect many niche solutions being developed and provided, that otherwise would not be marketable. For example when there is by itself a large enough group of customers but they are dispersed around the globe and in the traditionally way, it would not be profitable enough to market to them, they can now easily find you via an app in the iStore and viral promotion via social networks.

But also the services provided to larger organisations will benefit from this because in the end the existing offerings will become more efficient and cheaper. There will be economies of scale and there will be many standardised SaaS offerings that also large organisations can easily benefit from.

What will this lead to?
As a consequence, IT within organisations will change face (see also the article "The IT Jobs that Cloud Computing Will Create"). Technology itself will become more and more a utility and it will be much more about the business process, the feature or challenge that you want to resolve. The focus will be initially on managing the wild growth of various vendors and their services and to assure that these technologies will work in unison. I foresee changes in the IT roles required to manage all this.

Besides changing roles, I do foresee less technical roles within end-user organisations (I don't know whether the total number of IT staff will reduce, probably it will), a growth of roles with the service providers but this also will probably come with a concentration of those jobs in specific geographical locations.

For the techies amongst us there might be some hope that over time we will have so many mobile devices that the sheer amount of these might confuse the users and that they will need assistance with the use and configuration of those even if we can trust our vendors to improve the simplicity of it all. I just upgraded my OS on my iPhone but that process did not go that smooth and a less technical person probably would have given up or would have panicked. And the information that Apple provides at first hand about their iCloud offering is far from sufficient for me to decide whether I want to use it. I needed to go through some forums and blogs to get better insight.