There’s a notable excitement in the air about cloud computing. In fact, it's been there for a while, and you've almost certainly already noticed it. Or felt it yourself. Cloud computing - running applications over the internet in 'the cloud' - is, finally and irrefutably, on its way. The cloud is here, and it represents the holiest of Grails for businesses right now: that is, doing more with less. Those that are ready to make the most of the cloud, once it has changed from the much-hyped subject of conversation to a meaningful way of increasing innovation and productivity, will have a significant competitive advantage.
In a nutshell, cloud computing will save businesses a fortune. You could argue that it's all but free - hence that certain frisson in the atmosphere and the media.
It's also super-simple. Press a button and you've got more capacity. At least, that's the theory. But most enterprises won't set themselves up entirely in the cloud. Most of them, or their clients, have systems running on internal networks, and they are going to move portions of those systems to the cloud. Or they will expand into the cloud for new applications.
That means integration. Not only is there going to be a lot more traffic passing over the internet and through enterprises' network gateways, due to the fact everyone will be accessing applications running on the web, it will also be imperative to integrate your cloud applications to your enterprise and your enterprise applications to the cloud — again, all passing through your internet gateway, and all using relatively fat HTTP requests.
The million Gigabyte question is: "Is your network ready for this extra stress?"
The answer to this question will not just vary from one business to another; it will swing wildly from one polar extreme to another. Some enterprises will be network-optimised to the max, with intelligent networks underpinning powerful unified communications and virtualised data centres and applications. Good on them; they're the lucky few. Others will already be suffering from chronically poor network performance. They will have internal apps that run slowly now, applications that will slow to a crawl when accessed from the internet. As cloud computing takes off and then takes over, their network will, at best, creak under the strain. At worst, it'll collapse.
This is not to say these enterprises, or their IT staff or CIO, are at fault. The truth is that, for all that we've been excited about the cloud for some years now, the wider ICT industry has simply not used that time to properly warn and advise enterprises about the need to begin preparing their networks for the drastic change that’s on its way. We've not been vocal enough. Either that or our messaging has been too sales-focused and not sufficiently advisory.
The upshot is that there will be baffled or outright panicked looks on many IT managers' faces when next their office experiences a “slow internet day”. In the age of the cloud, they will no longer be talking about straightforward slow Google searches and interruptions to webcasts. Now, they'll have enterprise applications that aren’t working; and therefore an enterprise that is stuttering rather than running smoothly.
Networking problems are scary, because it takes a completely different skill set to identify and fix these problems than the skills typically developed as application developers. Cloud computing is an extremely cost-effective alternative to maintaining an expensive oversupply of in-house computing power. The challenge comes in ensuring that your network costs don't eat all of the savings. If enterprises are now actively adopting cloud computing, networking professionals have to change the way they think and how they do certain things in order to provide cloud computing networks.
There is no such thing as a "typical" application of computing. Companies use IT for highly distributed activities ranging from transaction processing to online retail; customer support to regulatory reporting. If these applications are hosted in the cloud, it becomes a prerequisite that cloud resources are linked to the company's own data centre resources for data access. It is also, clearly, necessary to provide user access to the applications in the cloud. Quite how this will affect the network will depend on three important considerations.
First up is the source of the data itself. Whether they are running in the cloud or not, applications always need data, and from where and how they get it is perhaps the most significant single factor in the network's impact on cloud computing cost and performance. Where there are large volumes of data involved, such as a large database - or several large databases, access to the data must be fast and reliable or the cloud application's 'runtime', its operational effectiveness, will be excessive. Enterprises either need to store the data in the cloud (which may present privacy concerns) or have a very fast and reliable network to support their cloud connections.
Next in line is the question of source data updates and backup. If the source data for the application is highly dynamic, or if it has to be backed up or synchronised with other company data, the link between the enterprise’s data and the application itself will need to be fast and efficient - very efficient. Where the data is more static, network requirements for maintaining the data will be less stringent and there may be little impact on a company's network.
The last of the three most critical network considerations for cloud computing is distribution of access. If the application is to be accessed from many locations in many countries, most of the access will probably have to come via the internet or an internet-based virtual private network. If access is primarily from a company's own facilities, then it may be necessary to connect the cloud computing resource to the company's own network.
Ultimately, the cloud, like the other long-overdue-and-now-finally-here ICT solution, unified communications, both relies on and puts extra stress on the network and its internet connections. As we move into an age where applications are hosted online and people and businesses communicate seamlessly by text, video and voice, the diversity and size of traffic on enterprise networks is going to expand significantly. Equally, the number of latency- and bandwidth-intensive applications associated with person-to-person, person-to-machine and machine-to-machine communications will increase.
With so many traffic sources, applications will become bogged down unless we start preparing networks pretty soon. Fortunately, this challenge is also the ideal opportunity for network intelligence to come into its own, making these applications run better, faster and more reliably. The latest generation of intelligent networks are designed to be self-configuring, self-healing, with automatic fault detection and correlation, self-protecting and, crucially, self-optimising.
Nobody should jump into cloud computing. Rather than a bold leap, it should be managed as a careful transition. A smart enterprise will trial out applications of cloud computing where network impact is minimal and gradually increase commitments to the cloud as experience of network optimisation develops. That way, the baffled and fearful expressions of IT managers across the globe will be kept to a minimum, while cloud-based cost savings are harnessed to the max.