15/10/2013 | Expert opinion - Karim Ogbi, Infrastructure & Facilities Management Business Unit Director, Hardis Group
Despite a certain maturity, many mid-sized companies and SMEs are still asking a number of questions about the private cloud. Preconceptions, real issues, points requiring vigilance: 6 key questions to help gain better understanding of this new way of consuming IT.
Is the cloud just about outsourcing your servers?
In outline, the cloud can be summarized as outsourcing IT resources from a third party. But the decision to "cloud" part of your infrastructure and its applications is a real IT project, which needs to be prepared beforehand.
First of all in terms of integration with other services or applications, internal or external to the company. Aside from some exceptions, an application requires a dialog between different bricks, in particular with a view to optimizing the use of enterprise data. Secondly in terms of the network. Outsourcing all or part of your information system leads to larger data exchanges with the outside world than before. It is therefore essential, in terms of throughput and associated costs, to provide adequate telecommunications infrastructure for optimal operation of remote applications.
Switching over to cloud environments is not something that can be improvised: it is better to start with some non-critical applications, in order to control risks when switching over "core business" applications.
Is there a right time to move to the cloud?
No more than with other IT projects, there is no "right time" to switch to the cloud, but there are factors that can trigger such a move. This is the case for example for a campaign to renew servers: is it better to invest in new machines, and upgrade the technical skills of the staff in charge of running them? Or to switch over to shared infrastructure, which by nature are constantly changing as technology changes?
The decision to switch to a private cloud can also be taken when an IS colleague (a systems engineer or infrastructure manager, for example) leaves the company, or when on-call time slots increase, or as part of a project to design or purchase a new application.
More generally, it is often a strategic decision to focus on the business that is behind the move and ensures that it is brought to completion: when general management considers that the added value of the information system and the IS team is its ability to assist business development rather than administer technical solutions and applications.
Does the cloud mean losing control of the information system?
The risk of losing control of your information system by choosing the cloud only exists if the internal IS department does not prepare for this new way of working. Instead of running its own IT infrastructure, its role will change, moving towards steering and controlling new services and the providers that offer them. This means that the cloud is not necessarily synonymous with downsizing the company's IT staff, but rather with changing the skills and tasks of the IS department staff. But the HR aspect is often neglected in such projects...
Is the cloud cheaper than in-house information systems?
Unlike most conventional facilities management contracts, the cloud delivers. First, it saves money: spending on initial investments are drastically reduced and distributed over the long term. For an SME, a private cloud is an excellent solution for accessing services and technologies that it would not have had the means to finance because they require significant investments in IT infrastructure.
Secondly, the cloud brings flexibility: a company can opt for a solution at time T, and change a few months later because its needs have changed. This a liberty that is hard to take when investments over several years have been incurred.
Finally, and as part of the high digitization of businesses and the increasing number of IT projects, the cloud enables the IS department to manage more projects, all things being equal from the HR standpoint, while benefiting from the expertise of the provider, which it would not necessarily have in-house.
And what about security?
While doubts about "pure" security (data theft or loss, protection against denial of service attacks, etc) appeared in the early years, the resources brought to bear by providers are generally much more powerful than those deployed by the clients themselves.
Today, the real questions to be asked are mainly about backup, data restoration and reversibility. In other words, the company's ability to recover data in order to reintegrate them into another solution, in the cloud or in-house. And these aspects must be contracted, and tested regularly.
What do you do with your in-house help desk?
This is a perfectly legitimate question, especially when the provider proposes his own hotline. But the answer is just as clear: keep it if you possibly can! For in functional terms, the in-house help desk will be able to provide level 1 support and respond to users who will find it reassuring to be able to contact someone working in the IS department, who knows the company and its business.
If this first level is not enough to solve the problem, the help desk can then qualify the malfunction and call on the provider. Appeals to the service provider will then be fewer and better qualified, and therefore less costly for the company.