Investment in datacentres is set to change, according to analysts. They predict that most sectors will invest more in innovation and upgrades relating to “edge” infrastructures rather than core facilities. Much of the expenditure will focus on automating links between core and edge facilities to reduce costs and improve efficiency.
Monitoring the state of the infrastructure in real time will therefore become a key issue. Aspects include IT and communication systems, power supply and cooling components and/or data related to the integrity of the site. In addition, solutions able to analyse and prevent spikes in demand for processing resources will be required to support operational efficiency.
For a datacentre, operational efficiency primarily means business continuity. The power supply, for example, must be continuously monitored to guarantee uninterrupted operation. A condition-based monitoring system can help to guarantee business continuity. It involves installing sensors that continuously detect aspects such as noise, vibration and the temperature of power supply devices. By collecting these data and the time series, it is possible to evaluate the residual lifetime of the facility and the probability of a fault. These systems can also allow manual input of data from inspections.
This is the first step in a more comprehensive predictive maintenance system. This involves configuring an algorithm for a single component that warns the plant operator when the component is approaching a potential failure.
Estimates suggest that by 2025, a fifth of global energy consumption will be related to datacentres. As a result, another increasingly important aspect of operational efficiency is energy efficiency.
The question is: how do we gather all these data from different sources, analyse them in real time and provide a clear output customised to the operator’s needs? RINA’s answer is RINACube, the platform where our monitoring systems are hosted. RINACube includes specialist software that can assess the energy efficiency of an asset such as a datacentre and support operational decisions to improve it. It can compare the asset’s performance with benchmarking data or with limits defined by the owner or manufacturer. Deviations that exceed a certain threshold trigger an alert, allowing timely intervention.
RINACube monitoring systems can be highly customised to the needs of operators, both in terms of input and output. The flexibility means the systems can be transferred to potentially any asset that requires monitoring. They can be used for a single ship or entire fleet, specific buildings or an extended area of real estate, a particular energy plant or a distribution network. The output is also highly customisable in terms of how the information is aggregated, compared and visualised, and alerts that can be activated.
Given the growth in automation in datacentres and other sectors, we expect the RINACube platform to strongly facilitate real-time monitoring for our clients in all industries worldwide.
find out more in RINA innovation + magazine