Cloud computing has become a marketing buzzword and is now used in many organizations. Yet the change to this system can represent a loss of control and the potential for disaster. Instead of fundamental shifts, IT decision makers want positive -- yet incremental -- change. They want to hear how the cloud can improve the way they work today, not radically alter it. In short, they want evolution, not revolution. The key is to balance the change and justify it on a case-to-case basis before making this change as a corporate decision.
Cloud computing is, in simplistic terms, the idea that you can offload your data storage and processing tasks to a very large set of computers, typically maintained by some large company such as Amazon. The novelty is that you abstract where the data is stored and which machine does the processing. None of this is really new -- even conceptually, as the "grid folks" have been pushing the 'compute-anywhere' vision for years. Some of the key unresolved issues are outlined here.
Sun -- when it was Sun -- claimed that it would eliminate all its data centers by 2015, but interestingly, they were moving into an open-source world where data security may not have been a major issue for them. Calling it "cloud" does not solve the issues that limit the growth of the grid. Data is valuable, and companies need to control where it is stored and who gets access. The proprietary algorithms being executed can be extremely valuable and are trade secrets in many industries.
It can take weeks or months to move a petabyte of data from one cloud to another -- depending on data transfer speeds, not to mention transfer fees. Why would you ever do it? There are many reasons that IT organizations might want to migrate, such as dissatisfaction with a cloud service provider, better alternatives, and changes in business or technology strategy. Any way you look at it, moving data is slow and expensive. The trend is not for this to improve (data sets are doubling annually in many industries, while CPU cycles are cheap and getting cheaper).
This is almost a non-issue today -- but it can completely shut down your business for a few hours, sometimes in peak time. For example, on June 30, 2012, Netflix and Instagram were down amid Amazon's Web services outage.
Loss of control
There is basically no insight into contingency procedures -- especially backup, restore and disaster recovery. The cloud service is a black box and users are tied to the online documentation explaining its procedures, not to mention that you are tied to the financial health of the cloud provider. In a May 2012 report, The National Institute of Standards and Technology (NIST) noted 23 “open issues” regarding cloud computing, including computing performance, cloud reliability, economic goals, compliance and information security.
Offline data synchronization
Networking issues are common in organizations. When this happens, the data will not synchronize with versions in cloud, and leads to versioning issues and chaos in group collaboration.
Sharing is not always good
There are several information security issues that remain open with the cloud, but the sharing of resources is a leading concern. "For [infrastructure as a service] clouds, different VMs may share hardware via a hypervisor; for [platform as a service], different processes may share an operating system and supporting data and networking services; for [software as a service], different consumers may share the same application or database," says the publication.
This is a huge issue for a few industries like pharmaceuticals. Some corporate data policies have stringent rules that PII (personally identifiable information) of patients cannot be stored in data centers outside the United States. The EU, for example, prohibits the cross-border transfer of PII data originating in the EU, unless the host country applies to certain EU regulations. In the case of data transfer from the EU to the U.S., the U.S. service provider must apply for the so-called "Safe Harbor Principle."
When data is on a cloud, anyone from any location can access that data from the cloud. The cloud does not differentiate between sensitive data and common data -- thus enabling anyone to access sensitive data. There is an inherent lack of data integrity in cloud computing. When you look at the cost of going across the network compared to the rest of your infrastructure, the realization is that local -- formerly known as distributed -- computing may still be the best route for most heavy data users.
In conclusion, you may consider using the cloud for specific elements of your business, such as putting some of your reporting in the cloud. If you are building a real-time reporting analysis engine, you may want to build that new business on cloud infrastructure to defray the startup costs.