Optimising data centre processes and reducing energy requirements is impossible without reliable information about all data centre components and the ways in which they are interlinked. This is true for IT systems as well as for power supplies, air conditioning systems, and building management systems. And there is another consideration: In the future, data centre infrastructure management (DCIM) must place greater emphasis on facilitating the provision of IT services.
There are a number of reasons to implement DCIM. This includes meeting stringent environmental regulations. For example, the European Energy Efficiency Directive (EED) aims to reduce CO2 emissions from the 1990 levels by 25 percent in 2020.
Additional factors that make DCIM indispensable are:
- Higher complexity due to the implementation of virtualisation and new technologies, such as cloud computing and app stores for mobile systems, and
- Pressure to rein in IT costs, especially in data centres. According to Gartner, IT managers worldwide can expect funding for new data centre systems to increase by only 3.7 percent in 2014, significantly less than for (mobile) devices (7 percent) and software (6 percent).
Using DCIM to monitor and analyse data centre components on an ongoing basis is the only way to identify scope for optimisation.
Core functions of existing DCIM solutions
The core functions of conventional DCIM platforms include capturing the components in the IT environment, such as servers, memory systems, switches, and UPSs, as well as the infrastructure components (power supplies, air conditioning systems, and building management systems).
In addition, they measure utilisation of systems, their power consumption, and the status of the air conditioning systems. This data is used for performing various tasks, such as:
- Managing components, including configuration settings,
- System, network, and infrastructure management,
- Capacity planning, simulating changes, determining long-term trends, and
- Automating work processes.
Data centre specialists ideally also have access to data allowing them to perform tasks, such as optimising rack and server workloads.
However, existing DCIM strategies have their weaknesses, one example being the fixation on IT and data centre infrastructure, which means that virtualisation of servers is not adequately addressed. The same is true for cloud computing, i.e., the use of "X-as-a-service" offerings (software, infrastructure, platform, business process as a service). This is problematic because IT departments then have no way of gathering valid data on the real costs of these cloud services and developing sustainable billing models. This affects not only cloud service providers, but also companies that deliver IT services to internal users via a private cloud.
Conventional DCIM solutions also fail to take sufficient account of business services, thereby preventing a comprehensive view of the data centre. The cost of providing an IT service via a specific physical or virtual server in terms of power, air conditioning, provision costs, etc., remains unclear.
Sign up for CIO Asia eNewsletters.