Evolving challenges for data centres


    MCP asked Schneider Electric’s Marc Garner: what are the current challenges facing data centres and which technologies are coming to the fore to solve them?

    For internet giants, colocation providers and many end-users, the key data centre challenges are around speed to market, scalability and energy efficiency, according to Marc Garner, vice-president of Schneider Electric’s IT Division for the UK. 

    “The current boom in data means that many of these companies cannot build fast enough – for colos it’s all about how fast they can get their data centres online. At a colocation level, the focus is on competitiveness. They need to differentiate their business from the next provider and they also need to get to market faster. 

    “Customers are now asking ‘can you get a data centre that is operational in 16 weeks?’ If they can achieve this in 15 weeks there is a good chance that they will win the business,” says Garner. 

    He points out that this is where prefabricated power, IT and cooling modules for data centres come in. Standardised, pre-assembled and pre-integrated modules can save deployment time and upfront cost, compared with traditional approaches, eliminating the need for custom engineering and considerable amounts of onsite work. 

    The modular nature of prefabricated solutions also enables rapid scaling and rightsizing to actual data centre loads. This, in combination with current power and cooling distribution technologies, is calculated to result in TCO savings of nearly 30% compared with a traditional data centre (findings in Schneider Electric White paper 164, TCO Analysis of a Traditional Data Center vs. a Scalable, Prefabricated Data Center).

    Other challenges are also emerging around the efficient cooling of today’s data centres. Largely driven by the recent explosion of IoT, the processing power needed for workloads at the edge and AI applications in data centres, which require advanced solutions to keep connected systems operating efficiently. 

    With a view to addressing this challenge, Schneider Electric recently teamed up with Avnet and Iceotope to jointly develop liquid cooling solutions for data centres. Schneider Electric’s preliminary analysis of a chassis-level immersion cooled solution, versus a traditional air-cooled solution, shows capex savings of 15% and energy savings of at least 10%, which leads to a 20-year TCO saving of more than 11%.

    In addition to these challenges, there is increasing pressure to tackle global carbon emissions and data centre energy consumption is coming under particular scrutiny.

    “Energy efficiency has become extremely important on an operational level, both financially and from a corporate social responsibility perspective, for the data centre sector. The more operationally efficient colocation providers can become, for example, the more cost competitive they can become for their clients,” Garner continues. 

    This is where the use of AI and intelligent software can make a real difference. Intelligent software solutions can provide key insights to help to drive efficiencies, bringing all the various technologies together and ensuring visibility across the data centre. Next-generation DCIM platforms are beginning to take advantage of AI technologies that will make management more predictive and automated. 

    Through adoption of next-generation DCIM architectures, facility managers can ensure maintenance upgrades and service schedules perform at optimum levels. Next-generation DCIM platforms, such as Schneider Electric’s EcoStruxure IT, are defined by five key attributes, which differentiate them from legacy data centre management systems (see box). 

    This level of visibility across critical infrastructure will be crucial for distributed IT environments. Network bandwidth and latency issues are driving the need for solutions at the local edge – outside the core data centre and closer to the processing application. 

    “Here, the challenges are around competency and skills sets – many businesses will not have an individual on site that will understand what the ‘flashing light’ means for their rack,” says Garner. “Remote systems administrators and in some cases, service partners or managed service providers (MSPs), will need to monitor changes within critical equipment such as uninterruptible power supplies so that system uptime can be maintained.” 

    This is where intelligent software, such as EcoStruxure, could hold the key. Data centre operators can set thresholds that generate alarms via text message, email or via the EcoStruxure IT Expert smartphone application when conditions such as temperature and humidity rise beyond acceptable levels. 

    Furthermore, the technology uses data analytics to monitor equipment in real-time and proactively mitigate unplanned downtime from anywhere, at any time and on any device. 

    In scenarios where multiple edge computing systems are being monitored, the ability to mass configure similar system characteristics (for example, rack locks, temperature thresholds) to similar devices, all at once, can generate significant time and cost savings. 

    Garner adds that the technology also “allows users to benchmark one of their edge sites or infrastructure solutions in one geography and compare it to another”. 

    “The heat in one rack may be twice as much as any other location and the software will enable proactive intervention, alerting the need to send an engineer to the site. If the technology can offer a return on investment of around two years and reduce the energy bill by 10-20%, this starts to provide a funding model for investment,” he comments.

    Resilience is also crucial, for edge computing sites, he points out. “If you look at the amount of data being captured in healthcare, for example (through patient records and clinical decision-making technologies), the risk associated with not investing and moving with the times is significant. There is an increasing need for edge IT applications within the healthcare setting and this needs to be managed effectively, through intelligent software platforms,” Garner asserts. 

    Another challenge facing edge facilities is a lack of physical security and resilience. Prefabricated and pre-engineered data centre technology can also help to overcome these issues. 

    The EcoStruxure Pod Data Centre, formerly known as HyperPod, for example, features integrated power, cooling, cabling, software management and containment – enabling racks of IT equipment to roll into place, similar to a docking station, without the complexity and time associated with traditional IT deployments. 

    There has been increasing interest in the EcoStruxure Pod Data Centre solution in the UK – particularly within the life sciences sector, according to Garner. “For life sciences customers, those that produce high volumes of critical data, the benefits have been significant in terms of speed of deployment and scalability. However, energy savings are also a key feature, since the pod architecture can also enable more efficient cooling. In some cases, clients are expected to see a return on investment in just a couple of years,” he says.

    There is also rapidly growing interest in edge solutions within the retail industry. To address the challenges in this sector, the company has recently introduced an industry-first EcoStruxure Micro Data Centre 6U Wall Mount, which allows large edge servers to be mounted securely in small spaces and announced partnership with Scale Computing, to deliver the ‘HC3 Edge for retail’ solution in EcoStruxure Micro Data Centres. 

    Scale Computing’s HC3 Edge for retail was developed to meet the variety of challenges retailers face when updating their edge and remote infrastructures. Shipped fully integrated and ready to deploy, this certified pre-integrated solution provides complete IT infrastructure within a secure, standalone enclosure for improving business’ uptime and resiliency, while providing protection of critical retail business applications. 

    Retailers can also leverage EcoStruxure Asset Advisor (Schneider Electric’s expert service bureau) for 24/7 remote monitoring and troubleshooting support. 

    “Today’s retailers need their in-store infrastructure to be highly secure and resilient. However, they also need to speed up deployment to the point where they can go from purchasing a site to becoming operational in just six weeks – standardisation and prefabricated technology are key to enabling retailers to achieve this,” Garner concludes.

    Next-generation DCIM 

    •  Relies on cloud technologies for ease of implementation, scalability, analytics and maintenance
    •  Connects to a data lake enabling insight and event prediction with AI
    •  Utilises mobile and web technologies and integrates with third party platforms
    •  Prioritises simplicity and intuitive user experiences in its design
    •  Serves as a compliance tool to identify and eliminate potential cyber security risks 



    Please enter your comment!
    Please enter your name here