Rack, Stack, Deploy

Hyperconverged infrastructure is essential when embracing the demands of the IoT. By Lee Griffiths, Infrastructure Solutions Manager, IT Division, APC by Schneider Electric.

The explosion in digital data requires, among many other things, an expanded vocabulary just to be able to describe it. Many of us familiar with computer systems as far back as the 1980s—or further—remember what a kilobyte is, in much the same way as we remember the fax machine. However, millenials may only have a vague recollection that such terms or technology ever existed or were relevant.

 

Now we must become as familiar with terms like the Zettabyte (that's 1x 1021byte) as we are with products like smart phones and self-driving cars—marvels that were science fiction not so long ago but which have become, or are rapidly becoming, commercially available realities. Industry analyst IDC predicts that by 2020 there will be 44Zbyte of data created and copied in Europe alone, based on the assumption that the amount will double every two years.

 

If that projection sounds like an echo of the famous Moore's Law of semiconductor growth, that the number of transistors able to be constructed on a single silicon chip will double every two years, it may be worth while pondering the effect on the computer industry and the world of that simple empirical observation of the 1960s. The ever-increasing capability of silicon chips, both to process information more quickly and store and retrieve it in vaster capacities, has underpinned our information society for decades.

 

More recently Cisco has estimated that within 5 years 50 billion devices and ‘things’ will be connected to the Internet. Digitalisation is driving the economy and according to IDC, by the year 2020 around 1.7 megabytes of new information will be created every second for each human living on the planet.  Incredible numbers, but when you look at new innovations such as Amazon go, the partially automated Grocery Store, you can see these predictions becoming a reality.  Think of the amount of IT Infrastructure that will be needed to support the supermarket industry of the near future if this new way of shopping takes off!

 

IDC's projections warn those depending on digital services that the amounts of data required are only going to increase for the foreseeable future. How does the data centre industry respond with the necessary capacity?

 Speed of response is a vital factor in the data centre industry. Indeed, the very rapid surge in creation of new data and the consequent demand for more data centres is causing the industry to diversify along two routes.

 

At one level, centralised data centres are becoming bigger, making huge data capacities available. At another, smaller data centres are moving to the edge of the network, bringing data closer to the point of consumption, simplifying Internet traffic and reducing network latency for applications such as video streaming - where speed of response is essential. The rapidly growing Internet of Things (IoT) phenomenon is also leading to a demand for smaller data centres distributed around the edge of the network. 

 

Collaboration enables innovation

 

The response of OEM’s in both the data centre and IT industries has been to realise that no one vendor can provide all tools necessary to deliver the types and variety of digital services required by todays rapidly growing businesses; that the necessary speed of response requires a collaborative approach between vendors, and their channel partners the systems integrators charged with designing and assembling facilities to customers' specific needs; and that standards and interoperability between various vendors' products are essential to such a collaborative approach.

 

The goal for many skilled systems integrators now is to be able to rack, stack, label and deploy a data centre of any size for any customer, tailored to their particular needs and rolled out in the quickest time frame possible.

 

At the most specialised space of the market this comprises localised Edge or micro data centres, delivered in a single rack enclosure with integrated power, uninterruptible power supply (UPS), power distribution, management software (DCIM), physical security, environmental monitoring and cooling. Such infrastructure can be assembled and deployed rapidly to support a self-contained, secure computing environment, in some cases in as little a timeframe as 2-3 weeks.

 

Further up the scale are centralised, hyperscale data centres with purpose-built computer Halls  comprising air-conditioning equipment, containment enclosures for Hot or Cold Aisle cooling configurations and the necessary power supply and networking infrastructure to drive a multitude of computer-equipment racks. Such installations need to be adaptable to accommodate rapid upgrading and/or scaling up or down of the amount of compute and storage capacity to respond to end-user needs.

 

Simplified integration

 

In either case, the ability to scale rapidly, to rack, stack, label and deploy the solution requires ever greater collaboration between vendors and convergence between their various product offerings, so that the time taken to build complete systems is as short as possible. In many cases Data Centre companies offering critical infrastructure solutions must work closely with IT Vendors who produce rack-mounted servers, storage arrays and networking equipment to ensure their products integrate seamlessly with each other.

 

Integration is absolutely essential for end-users looking to embrace digital transformation and expand their footprint rapidly. Solutions must be delivered ready to deploy, in excellent working condition and that requires both focused partnerships and the skills of highly specialised System Integrators who have become the go-to people in the converged and Hyperconverged infrastructure space. The magic it seems lies not within the individual pieces of IT and infrastructure equipment, but very much within the way the system is built, tested and deployed.

 

Additionally, such hardware must be guaranteed to work flawlessly with DCIM (Data Centre Infrastructure Management) and virtualisation software solutions that allow pools of processing or storage resources to be treated as individual isolated systems dedicated to particular customers or applications.

 

Collaboration; the key to hyperconvergence

 

By definition, converged infrasturucture enables the user to deploy four critical or core components of data centre technology - the compute, storage, networking and  virtualised server - within a single, secure, standardised, rack-based  solution. Hyperconverged infrastructure differentiates by utlising software to enable tighter integration between components and recognise them as a single stack, rather than as individual products.

 

In many of todays markets many businesses are adopting hyperconverged solutions as a more collaborative, forward-thinking and customisable approach for their data centre infrastructure requirements. It means that they can strategically hand pick the core components, which are in many cases used to expand footprint whilst providing both resiliency and connectivity at the Edge of the network. The real beauty of hyperconverged infrasturcture is that once a particluar solution is chosen, tested and deployed, it can be quickly standardised and replicated to provide faster scalability and reduced costs – both in CAPEX and OPEX.

 

A recent example of collaboration between Vendors is that between Schneider Electric and companies such as Cisco, HPE and Nutanix. In the former case the two companies have worked together to certify that Cisco's Unified Computing System (UCS) servers can be shipped already packaged in Schneider Electric's NetShelter racks and portfolio of localised Edge or micro data centre solutions. Nutanix, meanwhile, has certified that Schneider's PowerChute Network Shutdown power protection software will  work seamlessly with the  ESXi and Hyper-V  software used in the management of their own Hyperconverged systems.

 

In addition Schneider Electric has leveraged its Micro Data Center Xpress™ architecture in partnership with HPE on HPE Micro Datacenter, a collaboratively engineered converged infrastructure solution providing end-to-end IT infrastructure, networking, storage and management in a self-contained and easy-to-deploy architecture - ideal for distributed Edge Computing and IT environments.

 

Collaborations such as these and others help simplify the task of systems integrators as they specify and assemble bespoke data centre systems of all sizes. Giving them the peace of mind that key components of the overall systems they are tasked to build will work together seamlessly, reliably speed up the delivery time of new data centre deployments and allow their customers to rapidly  scale businesses as they seek out new markets. It is therefore of paramount importance that Edge data centre solutions work reliably, as promised from the moment they are connected.

 

The advance of semiconductor technology, guided by the road map established by Moore's Law, brought in the era of the PC, cellular phone networks and handheld technology. Now, in the era of Cloud Computing and the Internet of Things, the data centre – no matter the size -  provides the fundamental technological base that makes all other services possible.

 

The ability to rack, stack and deploy new IT resources quickly, efficiently and under the guarantee that they will perform to specification, will no doubt have a huge impact on how well companies succeed in the era of Edge Computing and the IoT.

 

By Krishna Sai, Senior VP of Technology and Engineering.
By Danny Lopez, CEO of Glasswall.
By Oz Olivo, VP, Product Management at Inrupt.
By Jason Beckett, Head of Technical Sales, Hitachi Vantara.
By Thomas Kiessling, CTO Siemens Smart Infrastructure & Gerhard Kress, SVP Xcelerator Portfolio...
By Dael Williamson, Chief Technology Officer EMEA at Databricks.
By Ramzi Charif, VP Technical Operations, EMEA, VIRTUS Data Centres.