Business Innovation: Optimising The Cloud Infrastructure

Cloud Infrastructure

Businesses are increasingly incorporating cloud infrastructures into their strategies. The cloud helps businesses realise real rewards, such as increased efficiency, flexibility and cost savings, but a very big hurdle is often overlooked – the network. Loss, latency and bandwidth limitations result in unreliable connections, and slow data transfers.

Poor network stability can not only increase organisations’ expenditure as they try to make up for poor connectivity by buying more wide area network (WAN) capacity or upgrading servers, but it can severely inhibit any innovative business projects using the cloud.

Cloud Computing Challenges

One way of understanding the challenges associated with the cloud is by comparing it to the weather. Just as the weather is unpredictable, cloud infrastructures also face unpredictable traffic demands and application performance challenges that can wreak havoc on networks, data centres, and remote offices all connecting to the cloud. Cloud computing involves, at the highest level, the delivery of hosted services via a shared WAN.

Irrespective of the type of cloud computing initiative, they all have one thing in common – data is centralised, while users are distributed. This can create extreme traffic peaks and troughs, which the network has to contend with – much like we have to when faced with harsh weather conditions.

This unpredictable environment can de-stabilise the cloud making it susceptible to the same WAN bandwidth, latency, and network quality changes that impact other enterprise applications. While we cannot prevent bad weather from happening, we have learned to prepare for it by designing buildings that structurally endure harsh weather, boarding up windows, sandbagging, and warning people to evacuate. In the IT and business world, we also need to prepare for the unexpected.

Three network elements that impact cloud connectivity are bandwidth, latency caused by distance and poor network quality caused by packet loss. The relationship between the three is complicated, with some having a greater impact in any given network environment.

Adding more bandwidth, for example, will not always make a difference to cloud projects, particularly those over long distances where latency is a major factor. Similarly, all the bandwidth in the world will not matter if packets are being dropped or delivered out of order due to congestion, as is often the case in MPLS and internet connections.

Wanting WANop

Deploying a predictable cloud service requires addressing the underlying network issues. Failing to do so will result in an environment plagued by issues which will only lead to performance and business benefits being compromised. To do this, organisations need to optimise the WAN, which can reduce over 90 percent of the traffic across the network and provide the scalability needed to support all current and emerging applications.

As virtualisation becomes a growing priority, we are seeing the data centre evolve, with virtual software increasingly replacing physical appliances. Virtual WAN optimisation is essential for the migration of large volumes of data as it not only saves on hardware, shipping and deployment costs, but it ensures that the data is easily and rapidly available.

With virtual WAN optimisation, businesses have the power and flexibility to build their cloud and data centre networks to withstand unexpected traffic congestion, reduce the bandwidth-consuming effects of redundant data, overcome latency challenges for long-distance users, and easily build-in “bursting” capabilities to address anticipated spikes in cloud-based application usage.

To do this, all three network challenges need to be addressed to optimise performance across a WAN, which will help guarantee a successful cloud investment. This can be achieved with three real-time optimisation techniques.

Latency mitigation using various protocol acceleration techniques overcome the impact of distance, Forward error correction (FEC) and packet-order correction (POC) overcome packet delivery issues in real-time, while quality of service (QoS) enables businesses to prioritise key traffic and ensure it gets allocated necessary resources. Finally, deduplication and compression maximise bandwidth utilisation.

A big benefit is that the investment made to improve cloud initiatives can be amortised across other enterprise applications. More data can ultimately be accessed and transferred between the cloud and the user in less time – making it even easier for organisations to enjoy the benefits moving and storing data in the cloud has to offer.

Finally, as the shift to the cloud can often change a company’s threat profile, businesses are under increasing pressure to secure critical communications and data transfers. As a result, not only will security policies need to be updated and evolved to accommodate these changes, but any WAN optimisation solution will need to address this so that data is accelerated in transit securely between data centres, remote offices and the cloud.

A fully equipped network enables companies to easily meet their business objectives, while allowing them to take advantage of their cloud investment. With cloud so often associated with forward thinking and business innovation, it is imperative that businesses ensure that they have the most reliable infrastructure in place to drive this. Fail to do this, however, and the cloud investment will either underperform or fail entirely – a risk that today’s businesses cannot afford if they wish to stay one step ahead of their competitors.

Dave Greenfield

Dave Greenfield has spent more than 20 years as a journalist and independent technology consultant. Today, he serves as a product marketing manager at Silver Peak. He has previously served as editor and blogger at a number of publications tracking WAN optimisation and virtualisation.