A Guide to Edge-Based Data Centre Deployment

Avatar photo

There is simply no question that our society is becoming ever more dependent on data usage. Enterprises need access to various forms of constantly updated information to help them achieve their business goals, while industrial sites must be able to analyse such information in order to make their operations more efficient. Also, a greater emphasis on cloud services is now being witnessed – with the increasing popularity of telecommuting, as well as collaboration between people in different geographic locations, only adding further to this. On top of all that, there are the changing demands of the general public to consider – with mobile subscribers wanting instant availability to an array of different services via their smartphones.

As a consequence of the activities just outlined, data centre facilities are destined to become even more valued in the years ahead. At the same time, it should be noted there is a shift underway in the topologies being employed here. Over the course of the next few years, a much larger proportion of computing and data storage resources will need to be situated at the edge – rather than everything tending to be centralised like it is now.

This is the motivation behind micro data centres – scaled-down versions of the larger facilities that we are more familiar with. These can be situated much closer to where the action is, so that edge computing tasks may be undertaken. As we will see, there are various operational advantages resulting from this.

The emergence of edge computing

There are several dynamics driving migration towards the edge. Among these are:

  • Mobile subscribers wanting to play immersive games or experience AR/VR apps (without lag issues).
  • Industrial systems needing to be able to deal with situations that could otherwise put staff in danger or have severe cost implications in as short a timeframe as possible.
  • The move to higher levels of vehicle autonomy in the years ahead (and the need to support safety-critical functions).

Analyst firm Gartner predicts that by 2025, at least 75% of enterprise-generated data will be created/processed outside of traditional centralised data centre or cloud operations – with a considerable amount of workloads being executed at the edge. Fuelled by this, the global edge computing market is going to experience a 38.9% compound annual growth rate (CAGR) between now and 2030, according to Grand View Research.

Through edge computing, it will be possible for a plethora of benefits to be derived across a broad range of different sectors. These include:

  • Placing greater autonomy where it is actually most needed, thereby accelerating responsiveness.
  • Reducing network congestion – as fewer data needs to be passed through network infrastructure back to a centralised point.
  • Heightened security – since fewer data transmissions over such networks will significantly lower vulnerability to malicious third-party attacks.
  • Greater scalability – with an edge-based approach making it easier and more cost-effective to add extra resources than having to do so at a large-scale data centre facility.

Overcoming the associated technical difficulties

Though this all appears to be very positive, having data centre operations situated at the edge will present challenges. Acute constraints will be placed upon deployed equipment that were simply not applicable to large, centralised sites (in terms of power consumption, available space, budget, etc.). There could also be limited site accessibility to factor in (impacting on maintenance work).

As with centralised facilities,edge-located data centres must have an assured supply of electricity, so that the integrity of the data that they store is fully assured. There are various issues that might potentially pose a threat here. These include power outages, surges on transmission lines and possibly even lightning strikes. The installation of an uninterruptible power supply (UPS) is therefore essential.

Next, an effective thermal management strategy has to be defined (the need for this being accentuated by the density with which hardware must be packed together in micro data centre installations). Depending on the amount of heat being generated, there are various, different methodologies that can be used here. Then there is the matter of having adequate data connectivity in place.

There is an undoubted need to keep both initial financial outlay involved and ongoing operational expenses down, as well as remaining aligned with environmental guidelines. To address increasing edge-based data centre deployment activity, Distrelec offers a wide variety of different hardware, software and test solutions from a broad array of relevant supply partners.

UPS Solutions

Before selecting a UPS for micro data centre installation, it is necessary to establish what the total data workload is going to be. This will mean the UPS may be correctly sized, with no shortage in the time it can support operation nor any excess when it comes to the cost.

Caption 1: Examples of Vertivℱ Edge UPS units

Available in mini-tower, 2U/3U rack/tower and 1U rack-mount configurations, with models that have 500W to 3kW power ratings, the Vertivℱ Edge UPS series is highly suited to decentralised data centre implementations. Each incorporating up to 10 power outlets, these units can deliver efficiency figures reaching 98% and have a 0.9 power factor.

APC’s Smart-UPSℱ units have the connectivity and remote monitoring functions needed to make them easy to manage from any location. Details on their status can be viewed via a secure web portal. All the various models within this family can deliver instant back-up battery power if an outage occurs, with pure sinewave outputs that mean sensitive equipment is not left vulnerable and heightened efficiency is assured.

With the capacity to be accommodated into either rack or tower implementations, Eaton’s 1kW 5PX Gen 2UPS units have a significantly larger output power than competing UPS products – making them able to support a larger number of servers. Energy Star 2.0-certified, they offer elevated levels of efficiency, thereby curbing electricity consumption and keeping down thermal management costs. Proprietary battery management technology comprises a three-stage charging process that helps to prolong battery lifespan.

Caption 2: 9PX UPS from Eaton

The company’s 9PX UPS features a double conversion topology. By leveraging it, the unit is able to constantly monitor power conditions, as well as regulating voltage and frequency. This and the 9SX model have load segment control mechanisms that allow them to shut down non-essential equipment, thereby extending the battery runtime available to critical items.

Caption 3: Vertiv’s Avocent¼ ACS8000

Edge server solutions

The AvocentÂź ACS8000 advanced console servers from Vertiv have all the attributes needed for secure, remote data centre management. These compact form factor servers, which run on the Linux operating system (OS), each make use of dual-core ARM Cortex-A9 processing technology, alongside 1GB of DDR3 RAM and 16GB of eMMC flash. A broad array of interfaces have been integrated into them. Among these are cellular (3G and 4G), USB, Gigabit Ethernet, SPC fibre and legacy serial connections (RS-232, RS-422 and RS-485).

Connectivity hardware for industrial connectivity

Edge computing will serve as the foundation for Industry 4.0 roll-out, and the arrival of more productive and efficient ‘smart factories’. Siemens’ SIMATIC ET200SP presents engineers with a compact high-performance conduit for passing data between distributed sensor devices and their assigned edge-based processing resource. It encompasses a multitude of different communication options and can take care of data pre-processing, before sending only the higher-level data back. Using this approach will alleviate the workload that the micro data centre is subject to – with more rapid reaction periods, plus quicker commissioning times.

Caption 4: The Siemens SIMATIC ET200SP

Industrial-grade data storage

Another essential aspect of edge-situated processing operations is having adequate data storage capacity to complement them. This can be attended to either by hard disk drives (HDDs) or solid-state memories.

Caption 5: A 20TB SkyHawkℱ HDD from Seagate

Possessing up to 20TB of storage capacity, Seagate’s SkyHawkℱ HDDs are intended for use in edge-centric artificial intelligence (AI) workloads, such as object/facial recognition from locally deployed surveillance camera footage. They have 6Gbit/s SATA interfaces and are able to deal with up to 64 HD video stream inputs. Their 2M hours MTBF means long-term operation can be depended upon.

The traditional 3.5-inch format WD Goldℱ enterprise-class HDDs from Western Digital feature the company’s proprietary HelioSeal¼ technology, which allows industry-leading storage capacities to be attained (reaching 22TB). Inclusion of vibration protection technology augments the reliability of these drives, with real-time monitoring helping to correct linear and rotational vibration issues.

Accompanying software

The Distrelec portfolio also includes Siemens SIMATIC software. This is optimised to support edge-based server infrastructure in uncompromising industrial settings – maximising operational effectiveness and energy efficiency, as well as safeguarding against cybersecurity breaches.

Network testing aspects

Assuring the integrity of both copper and optical fibre ethernet infrastructure employed in micro data centres is another important facet to consider, as any issues here could have major consequences. The NetAlly portfolio of test and analysis includes handheld testers for checking Gigabit Ethernet and PoE cabling performance within the shortest of periods.

For instance, NetAlly Portable Network Expert diagnoses any network issue, wired or wireless by enabling engineers and technicians to complete more tasks more efficiently with the help of a multi-technology. This all-in-one portable network analyzer allows users to deploy, maintain, and document their Wi-Fi, Bluetooth/BLE, and ethernet access networks.

Caption 6: EtherScope nXG Portable Network Analyser


It is clear that edge-based data centres are set to play a pivotal role in the digital transformation of many aspects of our society. Through these upshots like greater autonomy, augmented security and accelerated responsiveness will all be accomplished. Nevertheless, there will be acute pressures and resource constraints that these operations are going to be placed under. For this reason, sourcing optimised technology will be paramount.

Previous Post

AVAILABLE from DISTRELEC: The World’s First Oscilloscope with Update Rate of More than 4.5m Waveforms per Second

Next Post

Rugged Industrial Switches for Datacomms and Networking

Related Posts