When it comes to justifying investment in DCIM software for metering and monitoring in a data centre, we have seen a few approaches adopted by customers. The days of 5-year paybacks are long gone – now any investment made in DCIM must be capable of being recouped within 12-18 months of implementation to gain support from Finance.
You can picture the scene: a chilly data hall with, perhaps, 1-200 racks in it which are fitted with metered power strips and/or there might be branch circuit metering deployed in the main PDU panels; there are temperature and humidity sensors; UPS; CRAC; and so on. What is missing is an automated monitoring system reporting in real-time on all of these elements, leveraging the investment in all the infrastructure hardware. The challenge is justifying the investment in such a piece of software.
One obvious benchmark to measure against is the current annual energy bill for the data centre. The benefits of this are its simplicity and tangibility – it is clear and understood by facilities, IT and finance alike – and most organizations will have at least an idea of its scale. Next step is identifying and quantifying potential savings which can be enabled and realized through the deployment of DCIM monitoring.
The following areas are good places to start:
- Metering and monitoring. It is reasonable to expect savings of ~10% through changes made as a result of metering and monitoring. Having real-time visibility of consumption, instantaneously available through reports and graphs, gives insight where there was none before. The data gathered from metering, when turned into actionable information by a DCIM tool, can result in significant savings when acted upon.
- Raising the ambient temperature. Each 1 deg increase in an over-cooled data centre equates to 3-5% savings in cooling energy consumption. (Effective temperature monitoring must be implemented to mitigate against any unwanted effects of overheating). So, for example, by raising from 21 deg C to 24 deg C, it can be argued that savings of 15% on the current energy bill are possible.
- Efficient operations. If the manual processes of data gathering, data manipulation and analysis, and management reporting amount to 2 days per month then that represents one person month per annum. Automating the processes will save that time which represents an annual saving of perhaps 5% on an estimated fully loaded cost basis.
- Customer service. Automated data gathering, reporting and alerting enhances the data centre’s customer experience by enabling the rapid and accurate answering of customer information requests for energy data, offering the potential for customers to self-serve specific information through a portal, and increasing customer retention through providing an enhanced level of service. Allowing for a modest cost of 3 days per month on such activities translates into potential savings of 7.5%.
In this example, the potential savings described above add up to the equivalent of around 1/3rd of the annual power bill for the data centre. Deploying a DCIM software tool to make it happen should be achievable with a rapid implementation time and a sub-year payback. Compelling?