The impact of cloud computing and virtualisation on the network

Our networks are beginning to bow under the pressure of the colossal data volumes that industry, government and consumers are now handling and transmitting. In fact, IDC predicts that the world’s data doubles every two years, with 1.8 zettabytes of data created during 2011. Many providers have recently – and very publicly – felt the impact of severe outages, causing millions worldwide to go without access to services, and these companies are still feeling the repercussions today. Despite ploughing extensive budgets into business continuity to protect customers (as well as an organisation’s integrity) from server or data centre downtime, the network does not often get the attention it deserves.

Now is the time to really think about your organisation’s network and whether it is up to the job. Could your network let you down at the worst possible time, or is your business taking smart (and often simple) measures to mitigate disaster?

Ignorance is bliss

One of the primary challenges for technology-dependent organisations is that the network manager, or team, typically has little visibility with business managers. The network manager finds himself or herself in a bind as they are driven by sourcing and maintaining reliable, robust technology, but the board is driven by budgets and efficiencies. Senior management often doesn’t see the need to throw money at a problem that doesn’t appear to exist, until it is too late.

Let’s go back to basics for a moment. The fact is the network is a complex and often misunderstood business entity. ‘Network’ is often used as a catch-all term for two or three parallel networks – at least one for storage (reliability, guaranteed data integrity and non-blocking performance) and one for data (typically unpredictable bandwidth and frustrating levels of complexity). Separate switches, host bus adapters (HBAs), network interface cards (NICs) and cables are required for each network. This clutter causes day-to-day pressures on IT department time, budgets, power consumption and cooling.

The IT department is often charged with building their own network monitoring and diagnostics tools to claw back some form of network visibility. While they offer some form of check on switches, bandwidth capacity and the flow of data through the network, they rarely cover all requirements, nor provide rapid diagnosis before outages can occur.

Planning, testing and simulation building ensures that business applications are resilient against stress and aren’t close to falling over anytime soon. Network managers are concerned about what the network is doing, while not knowing, and not being able to do anything about it. Knowing about bottlenecks and where they are, whether the network is near its limits, and whether a component is near the end of its useful life, are key to keeping the network robust.

Prepare for change

Business change is also a key variable on the capability of the network – whether this is a large influx of new employees, a merger, or a new service or product reliant on delivery by the network. How can your organisation account for business change if you do not know that it is a day away from imploding? The network should be resilient enough to cope with today’s demands, as well as those of the future. This is particularly true for organisations that rely on the network to deliver products and services to customers.

Systems upgrades, new business applications and anything else that sucks up network bandwidth should be prepared for. Changes can cascade throughout the business, and these situations are difficult to work back from.

A cloudy outlook

Given the widespread adoption of cloud computing, enterprise mobility and virtualisation, it is important to understand if the network is ready to support the planned growth of the business and respond quickly to the unexpected opportunity without time-consuming and expensive architectural changes. Business managers also need to control levels of customer expectation and, with any new network model, manage user education.

While cloud computing and virtualisation deliver considerable efficiencies, this has only added to the demands imposed on the corporate network infrastructure. Both clearly mean more traffic travelling across corporate networks and beyond the firewall: while virtualisation relies on one server to host multiple virtual machines and keep up with demand during peak times, the cloud has its distributed applications, residing on different machines and all performing different parts of the vital business process.

“We need double capacity for tomorrow’s launch …”

The business case for the network does not end here; business leaders need to recognise it for the foundation on which all progressive businesses depend, and its role in underpinning all future business growth. In today’s fast-paced business environment, business leaders often need to respond quickly and therefore having visibility of what a business has is crucial. Being able to flex and react to change without affecting performance – whether it is to increase capacity within 24 hours or simply add a few more users – will give an organisation a core edge on competitors. It is for this reason that technologies like Ethernet Fabric are making significant in-roads in achieving converged, highly reliable, robust and resilient, low-latency networks.

Keeping well-maintained and monitored networks that deliver continued resilience if something does go wrong means a certain amount of trust needs to be imparted onto the network manager, so that budget is not released at the last minute, by which time it may be too late. Visibility (of the network team and the network itself), preparation and planning and keeping a firm grip on projects, upgrades and applications can ensure the network is fit for the future and cemented on the corporate agenda.