Virtualisation in the data centre has driven up utilisation, consolidated IT estate and significantly boosted efficiency, all of which are excellent achievements in deploying technology innovations to deliver business benefits from IT.
While the advantages of virtualisation are clear, it does not come without its challenges, some of which are lying in wait for unsuspecting IT Directors. The ease of provisioning virtual servers can lead to virtual sprawl adding complexity to ongoing management. Also, with the computing workload server hardware now shoulders, the increasingly pressing challenge will be maintaining system performance levels from over-worked machines.
Legacy hardware is not built for virtualised data centres
At the heart of the issue is this: legacy systems simply aren’t designed to handle the increased I/O demands that virtualisation places on server CPU and memory. Whilst server virtualisation allows decision makers to ‘do more with less’, the flip side of this is that the server itself is being asked to take on increased workload, often up to ten times the work of one server, with exactly the same hardware configuration and capability. It’s not difficult to see that this equation does not balance.
Often virtual servers become overwhelmed, leading to IT service disruption and slow running systems. Having invested in virtualisation however, IT professionals are still mandated to deliver the promised increases in productivity to management. With conflicting needs to hold productivity steady and data centre infrastructure creaking at the seams, the temptation is to find a fix and take the pain away. Provisioning new physical servers to add more computing power, or diverting traffic to a different virtual cluster applies the necessary band-aid, but it doesn’t solve the underlying hardware issue exposed by virtualisation.
Virtualising the data centre aims to reduce the burden of management and reduce complexity; a burgeoning server estate, especially when it is virtualised requires more, not less care and attention which little by little adds to the opex budget line. Essentially, working around virtual machine bottlenecks stores up issues for the future and sets the business on the road to undermining the benefits achieved through virtualisation.
The time has come for IT Directors to start solving data centre issues and not just fixing them.
Integrated infrastructure stacks are purpose built for virtual data centres
Virtualisation, just for the record, is still a good thing. The way to maximise the business and IT benefits a virtual data centre can deliver is to give it a fighting chance by running it on hardware that can keep pace. The good news is that hardware vendors are onto this challenge and have taken steps to design new solutions specifically to accommodate the needs of today’s virtual data centre. Integrated infrastructure stacks hold the key to simplifying complexity and continuing to drive productivity, efficiency and performance.
VblockTM infrastructure packages from VCE, a coalition of Cisco, EMC and VMware, is the first example of an integrated infrastructure stack that can be purchased off the shelf as a complete hardware solution. Built on Cisco’s Unified Computing System (UCS), the Vblock solution combines best-of-breed storage hardware from EMC and virtualisation hypervisor from VMware into an integrated physical cabinet ready for the data centre. Cable management becomes a walk in the park with internal circuitry, eliminating the plethora of cabling required to connect traditional rack mounted servers to each other and to storage, replaced instead by just two cables for networking and power connectivity. As far as simplicity is concerned, the integrated infrastructure stack solves a multitude of management bear traps.
The leading waste management company in Germany’s Rhine-Main region, FES, is a great example of how integrated infrastructure can deliver benefits quickly over ‘traditional’ data centre design. It has driven 40% efficiency savings in power and cooling costs by refreshing its data centre infrastructure with the Vblock solution, which in addition to cost savings and simplified day to day management, has also eliminated the need for the company to install a major new air conditioning system.
A realistic foundation for public and private clouds
With the arrival of integrated infrastructure packages, the ability to transition to private or public cloud-based architecture becomes less risky and simplified. Commercially too, the models for both capital investment or for operational spend make sense for certain business cases.
For service providers or enterprise businesses that need cloud infrastructure, the presence of different vendors throughout the data centre has hampered progress through compatibility issues, risk and inaccurate ROI modelling. Building cloud infrastructure around the Vblock solution is as simple as this sentence. It’s already tested, it already works in an on-demand environment and the performance you can expect is documented alongside the cost.
Risk therefore falls significantly when investing in integrated hardware. IT Directors and CFO’s know exactly what they are buying and exactly what it will deliver for the business. It’s also possible to try before you buy. For instance, our Solutions Centres in the UK, France and Germany are available to customers to test proof of concept for private cloud and to measure for themselves the performance of the Vblock solution in combination with their existing IT estate.
There is also no sting in the tail from ongoing monitoring and operation of the Vblock integrated infrastructure. Ensuring best performance and proactive troubleshooting doesn’t need an army of administrators to be trained; it can be managed remotely, and delivered to users as a managed service such as MTI Care. With enterprise businesses and service providers keen to save money, the ability to outsource simple infrastructure management frees up their resource for more complex internal needs.
Refresh mind-sets for the refresh cycle
In the IT world, innovation happens often and quickly. Virtualisation revolutionised the data centre and paved the way for the utopia of cloud computing to deliver IT as a service to the business. As with all new technologies, there are bumps in the road, and whether we like it or not, the benefits of virtualisation will diminish over time. Performance will be more difficult to achieve without more servers, complexity of management will push up resource costs and ROI will continue to fall.
As hardware refresh cycles come around, we urge IT decision makers to think hard about how they can best maintain the benefits they have experienced and the goodwill they have generated by virtualising their environment. Hardware solutions that solve the challenge of virtualisation already exist; integrated infrastructure is an innovation that will allow businesses to future proof the ability of their IT systems to deliver quality service while avoiding risk and unnecessary investment.
 ShareThis
Tags: Virtualization |
