As a new year dawns, it's time to resolve to face up to so-called new IT trends and recognise that they are often merely marketing terms repackaged for the new generation and are far from new. They are merely re-inventions or groupings of services with a different name. At the end of May in a public meeting in Boston, Vineet Nayar, CEO at HCL Technologies, declared that the "cloud is bullsh*t", and I for one agree with him.
Vineet said, "My view on cloud is that I always look for disruptive technologies that redefine the way the business gets run. If there is a disruptive technology out there that redefines business I am for it. If there is no underlying technology there, and it is just repackaging of a commercial solution, then I do not call it a business trend. I call it hype."
Where's the underlying technology?
The key question to ask is, is there a key underlying technology or not? There is not. There are many forms of virtualisation, establishing shared services, "everything on demand" claims and being more effective and leaving spare capacity to be sold on to others. Weren't bureaux set up in the 1970s and 1980s to achieve the same things; wasn't workload partitioning on hard drives and partial software licensing to achieve lower cost and flexibility to clients? What is so different for "pay as you go" clients? Previous attempts, such as "IBM on Demand", were the same service but it never really took off.
Early adopters do sometimes pay the price to get a better return on investment (ROI) and to use enterprise shared service that they would not get otherwise. What shared services would those be then? Can you name any?
How is it that the biggest service providers in the world cannot provide a single simple agreed definition of what "cloud" is? Can we really look forward to consumption-based IT services for measured, sophisticated and complex organisations?
Can we really look forward to benefits such as:
- Cost containment - we need to develop total cost of ownership models to understand to see whether cloud can offer enterprise sustainable savings over time.
- Business knowledge and corporate memory - hard to retain as part of a true cloud offering. No transfer of assets, includes personnel. Bring back in-house - how would that work?
- Immediacy - act today and have the system working inside of a week.
- Scalability - allowing for real success (huge increases in capacity) and failure (no need to honour capacity expectations) and without expending undue sums to keep the capacity sitting around. Is this practical for suppliers who must invest upfront without guarantees of returns?
- Resiliency - mirrored solutions for disaster scenarios and load-balancing traffic.
- Availability - bandwidth in all senses. Without it, service delays could be disastrous as there is little back-up once our workload has been virtualised.
Risks and concerns are well documented in terms of security, trust in the longevity of the supplier, dynamic accountability of where the data actually is, compromises in confidentiality, disaster recovery of data spread all over the planet, etc.
The biggest risks come from legislation, not further legislation but existing legislation. For example, the Patriot Act that extends to every US company to report any and all transactions than breach the "norm" - norm not very well defined but large transactions, high value transactions, transactions undertaken in countries not considered "suitable" by the US authorities.
EMC, IBM, HP, CSC, Cisco et al could be part of the cloud and therefore they are obliged to report and to not restrict US authorities in their duties - with no obligations to their client to report Patriot Act activities. Still happy?
Private cloud public cloud - does the Patriot Act apply to private cloud? Yes if it is run by US-based service companies.
So back to the main point - is there a fundamental underlying technology that I can get anything unique from the cloud that I cannot from other outsourced services or old fashioned hardware/software ownership? No
The only two questions you now need to ask now are: is the lower cost versus security and other concerns worth it and, the question few think about, once totally committed and virtual, can we ever bring these services back under our direct control?