Basket £ 0.00 (0 items)
You are here: HomeArticle › The journey from grid to cloud computing

The journey from grid to cloud computing

Randy Clark of Platform Computing tracks the evolution of cloud enabling software from its origins in grid and high performance computing.

The buzz surrounding Cloud Computing has reached an all time high thanks to intense media interest and the marketing efforts of big brands such as Amazon and Google. According to IDC, growth in Cloud Computing is set to triple to £29.5 billion in the next three years1. Although there remains confusion over what is meant by 'cloud', Gartner's definition of Cloud Computing as a system where massively scalable IT capabilities are delivered as a service2 is accepted by most. IT-as-a-Service offers huge advantages to enterprise-level customers that want the option of significant computing power on tap without having to buy expensive supercomputers. The ability to dynamically grow or shrink compute power on demand by paying for it on a usage basis is particularly attractive given the current economic climate and pressures on organisations to do more with less. Yet despite all the talk around Cloud Computing and the benefits it offers, the technology has not arrived like a bolt from the blue. Instead, it is the result of an evolutionary process that started over 20 years ago, in part with technology known as grid computing, and there is still some way to go before we see widespread adoption of cloud.

The evolution of grid

The term grid computing first emerged in the mid-1980s after scientific organisations and government laboratories began to require greater computational power. The technology allows users to run complex, often process or data intensive applications on clusters of commodity servers. It offers organisations a greater level of flexibility and reliability by creating a system of shared resources that are able to compute large or complex tasks in a fraction of the time that a single computer would take by distributing the calculations across hundreds or thousands of CPUs. From humble beginnings in research labs, the technology accounting for £8.1 billion in sales in 20073 is now widely adopted for commercial use across all industry sectors including financial services, manufacturing, life sciences, oil and gas, government and education. Whether it be F1 teams using the technology to increase the amount of aerodynamic testing they are able to complete or pharmaceutical companies using it to go to market with new drugs faster than before, the benefits of harnessing the power of shared IT resources are now widely recognised.

In its infancy organisations adopted a departmental approach to sharing their resources but this has changed in recent years with enterprises taking more of a holistic view of their IT infrastructure. This approach involves identifying the processing power across an organisation and pooling it so it can be deployed where it is needed most. The set-up is often referred to as a shared resource or utility computing model and involves separate lines of business paying for usage, rather than ownership, to access computing resources for their applications. This stage has required a significant philosophical shift as users are required to share resources that they have traditionally always owned and controlled. IT departments have also had to adapt as they extend beyond just running the corporate network and adopt the responsibilities of a running computing service. Despite this, according to IDC, sharing is no longer seen as a dirty word and taking a utility model and applying it across an organisation is now seen as prudent4.

Issues impacting the adoption of cloud

It's not surprising that increasing numbers of organisations today are looking forward to the next stage - Cloud Computing. The concept of using the web as a means of delivering a service has huge benefits in terms of scalability and cost reduction but organisations have their concerns. In a recent survey of senior IT executives from leading financial services firms we found over half didn't think that 2009 would be the year that Cloud Computing would take off. Many cited security as a top concern inhibiting adoption. Although hosting data externally should hopefully result in avoiding the large number of high profile data breaches that were widely reported in 2009 by reducing the need for data to be stored on physical devices, it does raise issues regarding access. With so many people accessing data beyond the corporate network, organisations need to be assured that their sensitive data will remain secure and out of the hands of those with malicious intent. A second major issue is that of reliability. Like any Internet service, clouds can be subject to failure and subject to denial-of-service attacks. Although these threats can be addressed by working with multiple Cloud Computing service vendors, it is likely to be some time before organisations have the confidence to host all of their data outside the protection of the corporate network.

The rise of internal clouds

This reluctance does not mean that Cloud Computing will be ignored. Instead, what we will increasingly see is the creation of 'internal clouds'. 'Internal clouds' apply the concepts of Cloud Computing such as computing power available on tap and scalability but give organisations control over security, access and service levels. This prediction is supported by a recent global survey of 2,600 large and small enterprises, conducted by Forrester Research5, which found while more than half of firms are virtualising some of their servers, just five per cent are using hosted virtual servers. Clearly, enterprises are interested in the benefits the cloud can offer but first want to experiment within the safety of their own network. Once they are comfortable with the technology they will look to investigate how savings can be made by working with reliable third party suppliers of cloud services. Those feeling most comfortable will look to host some of their non-business critical applications in the cloud, while retaining sensitive data and applications internally.

The future of Cloud Computing

The concept of Cloud Computing may be a new term but it is the logical endpoint for the convergence of grids, virtualisation and automation. Many firms have taken the first steps towards the adoption of Cloud Computing, perhaps by employing a Software-as-a-Service (SaaS) solution, or leasing a few compute cycles from third party providers. Though issues such as reliability and security are legitimate concerns early adopters are learning how to overcome these barriers. These organisations are deploying internal clouds to share the IT resources they have available in house, as a step towards embracing a full scale adoption of Cloud Computing. Making this step requires enterprises to partner with experienced software vendors and deploy management tools that connect IT resources to workload demand. At a time when doing more with less is more important than ever before, expect to see demand for these solutions to rise.

References

  1. http://www.ihotdesk.com/article/19060314/Cloud-computing-will-reach-$42bn-by-2012,-IDC-claims
  2. http://news.cnet.com/8301-10787_3-10196859-60.html
  3. IDC, 'HPC Management Software: Reducing the Complexity of HPC Cluster and Grid Resources', 2007, p.16.
  4. http://www.platform.com/newsroom/Platform-on-track-Jan13.pdf
  5. http://www.vnunet.com/computing/news/2237868/firms-say-virtualisation-cloud

(ITadviser, Issue 59, Autumn 2009)

 

Contact

For more information about The National Computing Centre and our services, please contact us at the details below:

Email: info@ncc.co.uk
Telephone: +44 (0)870 908 8767
Fax: +44 (0)870 134 0931

Click here for more contact information


TwitterFollow us on Twitter
Linked InJoin our LinkedIn Group
FBLike us on Facebook

 

Management Guidelines

NCC Guidelines Vol 5 No 1

more in Management Guidelines

 

Professional Development

Cloud Computing

more in Professional Development

 

Analyst Digest

September 2016 Bulletin published

more in Analyst Digest