Data Centre Monitoring
Data Centre Monitoring
In land and resource scarce Singapore, going green when it comes to energy hungry data centres is not so much about the use of renewable power – since there is very little available, but more about the judicious use of energy. This could range from avoiding excessive use of energy, to finding ways to reduce usage without impacting performance.
This can be problematic for legacy data centres, which were designed more like black boxes to reliably deliver power and cooling to the servers and IT equipment housed within them. While DCIM (Data Centre Infrastructure Management) tools will certainly help, they are not always present and their effectiveness can vary significantly depending on how deeply they are integrated.
Fortunately, the energy situation is reversed where next-gen data centres are concerned. Indeed, this is where its comprehensive ability to monitor and control the data centre environment shines. These capabilities play an important role to help set the stage for a green data centre, whether by identifying wastages, or contributing towards the implementation and validation of new energy strategies.
As mentioned in “Understanding the 2.0 Data Centre”, the software defined data centre offers extensive control over all aspects of its operations, with very fine granularity over these controls. This means that it is possible to access a multitude of statistics such as temperature reading down to the rack level, as well as minute changes in the status of servers to better understand workloads and their corresponding power and cooling demands.
Indeed, the IO.OS platform was specifically designed to intelligently manage the entire infrastructure within the data centre, and offers real-time visibility including predefined alarms, temperature, humidity and air handling, among others. All these are consolidated and presented within a single pane of glass, and can even be accessed from a mobile app.
Moreover, this plethora of data is captured and archived, offering administrators the ability to review them retrospectively to better understand the inner workings of the data centre. This is important as servers are constantly added or replaced within data centres, which means that its energy footprint will not stay static.
Taking this a step further, it is possible to leverage this wealth of data to proactively tweak energy consumption throughout the day, as opposed to reacting to situations as they happen. For example, historical temperature data can be used to identify hotspots and have cooling autonomously adjusted to eradicate them. This ensures that the data centre stays cool without having to overprovision cooling.
At the macro level, this strategy could be extrapolated over the entire facility with the application of analytics to closely match cooling output to the heat produced by the hardware. Waste is dramatically reduced by replacing an estimated – or even arbitrary – configuration with one that is based on data.
While the above suggestions can also be implemented within modern data centres that are equipped with the right tools, next-gen data centres are in a unique position to effectively address the topic of efficient energy usage by means of their latent capabilities.