When IT departments are looking to increase efficiency, the data centre should be the first port of call. It’s an area where big gains can be made, especially if a data centre has been added to over time to meet current needs, or hasn’t been reviewed or audited. These days, increasingly complex demands placed on the data centre can mean that some assets are underutilised, others are redundant and the whole centre may be using more power than necessary. But maximising efficiency in the data centre is possible by following these three simple principles.
If you’ve never audited your data centre assets, make time to do so, and continue to review on a planned, regular basis. Start with the outside and tidy up. Simple tasks like checking stacks, racks and cables can identify little issues that, once rectified, help the centre run more smoothly. It’s also a common problem to have servers sitting idle or being underutilised due to a lack of planning or internal knowledge about function not being documented. Other servers may be past their use-by date. Older hardware can be slower and harder to cool, meaning they cost more to run. Replace old hardware with more energy efficient options that draw less power, and remember that the cheaper model to purchase may cost more to run in the long term: compare kW power usage and consider this over the life of the hardware. Power usage effectiveness (PUE) ratings help to measure how efficiently a data centre uses energy, so consider these when purchasing new hardware.
The business data centre consumes a huge portion of a business’s power usage, so finding ways to reduce power consumption is critical. If you don’t know how much power your data centre is pulling, you can’t measure any reduction, so make sure the IT department sees the energy bill and takes ownership of power usage. Use it as a benchmarking tool to monitor improvements. A major plus of this approach is a reduction in power usage equals a direct reduction in cash output and a better bottom line. Workload also determines the amount of power a data centre consumes. Review workloads and virtualise where possible, consolidate virtual machines, and get rid of unnecessary tasks.
The other major cost contributor is data centre cooling. Clever management of the data centre cooling system can have a big impact on data centre efficiency, and the overall cooling budget of the office space. Air economisers use outside air to cool the data centre, reducing the need for air conditioning. Likewise, isolating heat generating equipment and siphoning the heat out of the data centre, potentially into heating another part of the building, is a greener solution. If this isn’t possible, periodically varying the speed of the air conditioning system can help optimise its usage and reduce costs.
Finally, smaller can be better. There is no point paying to run equipment you don’t need. Consolidate servers and virtualise whatever you can, and if it’s not possible with the current equipment, consider upgrading to a contemporary system that allows for greater flexibility in data centre design.
Hyperconverged infrastructure, such as Cisco’s HyperFlex, uses a software-defined architecture to virtualise a traditional hardware-defined system, which reduces complexity and increases operational efficiency.
Find out more about how HyperFlex can revolutionise your data centre.
See the benefits of Cisco HyperFlex Systems in this special TechWiseTV broadcast.
See what’s new with our hyperconverged infrastructure built on Cisco Unified Computing System (UCS). Learn how we can develop, sell, and deliver together.