Data centre virtualisation and DCIM - a holistic approach
By Andrew Sylvester, Data Centre Software Manager, Schneider Electric, IT Business
Wednesday, 22 October, 2014
With energy usage and costs continuing to rise, expect to see businesses moving towards a more holistic and end-to-end approach to data centre strategy.
For the data centre industry, virtualisation lowers facility costs, simplifies administration, decreases asset management requirements and, largely, improves energy efficiency. At a time of rising energy costs and continual pressure for managers to increase data centre sophistication, virtualisation can improve the overall data centre performance and reliability.
In our experience, many IT and data centre managers only consider two layers to their data centre virtualisation strategy - IT infrastructure (servers, network and storage) and software. Often neglected is the third layer - the impact virtualisation can have on data centre physical infrastructure or DCPI.
Addressing DCPI, including cooling, monitoring and power, is vital in optimising virtualisation performance and in facing many of the challenges virtualisation can bring. Overlooking the holistic impact of virtualisation can result in compromised availability and often an economic loss.
Challenges to the virtualised data centre
Heat removal is an immediate challenge in data centre virtualisation. Virtualisation by its very nature drives larger capacity for data storage and helps data centres achieve higher utilisation rates. As a result, hosts draw more power and create more heat.
Managers also need to consider that when there is a reduction in IT load, with no change in the DCPI, data centre power utilisation effectiveness (PUE) will worsen, despite energy use decreasing. This is because although virtualisation increases the IT efficiency, it decreases the electrical efficiency of the overall system and causes excessive ongoing electrical consumption.
Most users often forget that there are fixed losses in the power and cooling systems that exist whether the IT load is present or not, and that these losses are proportional to the overall power rating of the system. These fixed losses are the dominant form of DCPI electrical consumption in most installations.
Virtualisation can also cause IT loads to vary in location and time. One of the advantages of virtualisation is the ability to move load when needed; however, shifting virtual loads combined with high density can increase the risk of downtime and unused or stranded capacity.
Seeing the complete picture
Data centre infrastructure management (DCIM) software provides a complete picture of the health of a data centre and is designed to identify and resolve virtualisation issues. It increases understanding of capacity and utilisation in the environment through monitoring and controlling DCPI in real time, based on changing loads.
DCIM can also help identify unused or stranded capacity by coordinating assets more efficiently and conducting predictive analysis of what will happen to the physical infrastructure before loads are moved.
Modular, scalable power and cooling solutions can also enable right-sizing to match the data centre’s consolidated or growing IT loads. Close-coupled cooling also adjusts to migrating IT loads in real time.
Cutting through the DCIM noise
While there are a lot of organisations offering DCIM, many are not able to provide the benefits that a full suite of solutions can provide. DCIM is not about point-solution products such as cable management and IT monitoring software that monitors only singular aspects of data centre operations - proper DCIM solutions cover the full suite of monitoring, automation, management and analytical capabilities.
The confusion on DCIM solutions can be compared to the analogy between the dashboard of a 1967 VW Beetle and a Toyota Prius. A 1967 VW Beetle only has information like speed and fuel, while the Toyota Prius can provide information such as how many kilometres can be done on the current tank and when an oil change is needed. Point-solutions run on device-level monitoring similar to the VW beetle. By contrast a full suite of DCIM solutions marks a fundamental change towards contextually aware monitoring - similar to what’s found in a Toyota Prius.
For organisations looking to take advantage of DCIM solutions, it’s important to start by determining which areas need to be focused on - availability management, capacity management, change management, solutions optimisation and so on.
Organisations should then look at modelling the outcomes of implementing DCIM before deploying it across the system, as this will ensure they have full insight into the benefits and potential ROI. Modelling is particularly relevant for managers who have limited visibility of their IT and data centre equipment and its performance, as DCIM can help reduce these concerns by providing a greater overview of the data centre operation.
When considering DCIM solutions, organisations should also take a combined view of data centre, facilities and IT ecosystems - this approach is called a ‘data centre management system’. This view considers organisational relationships and the impact of how people work and the processes they use across the organisation.
When properly deployed, DCIM provides invaluable insights into key cost areas within the data centre, providing information to managers on operational performance and day-to-day activities. The end result is that DCIM enables executives to make better decisions and manage IT operations more effectively. This results in improved management of resources in key areas such as power, cooling, space capacity management, energy cost management and business value.
In 2013, we saw many businesses taking on server virtualisation and data centre management. In 2014, with energy usage and costs continuing to rise, we expect to see businesses moving towards a more holistic and end-to-end approach to data centre strategy. Having a comprehensive energy management program such as DCIM will work to reduce increased energy cost and drive improved ROI.
Financial institutions have pushed public cloud adoption to the top of their priority list but...
For all the benefits the metaverse offers, it also introduces some tough challenges for businesses.
As part of the continuous evolution of SaaS solutions, both AI and machine learning are quickly...