Subscribe

Analytics in the era of hyperscale

Joanne Carew
By Joanne Carew, ITWeb Cape-based contributor.
Johannesburg, 07 May 2019
Sven Blom, Teraco.
Sven Blom, Teraco.

Today, businesses have to move faster, be more connected and operate in a more complex environment than ever before. Some may see this as a challenge, but it also empowers companies to reach levels that simply weren't possible in the past, says Bill Hoggarth, enterprise information management business unit manager for Gauteng at Datacentrix.

What do these new possibilities look like? Well, we live in a world of automated, digital connectivity - with all technologies that fall within this brave new world converging into one melting pot. The key benefits of our current circumstance lie in our ability to be more agile, more relevant and more sustainable as a business. Not only are these technologies creating a new breed of business model, they're also opening up an enormous amount of insight into how companies interact with their competitors, suppliers, partners and customers.

"Ignoring the hyperscale trend would be similar to being in denial about electricity replacing steam or cars replacing horses," says Nicolas Blank, Group CEO at NBConsult. He points out that handling big data analytics in the cloud actually requires surprisingly little in terms of changing disciplines and skills. Security is still security; how it manifests and how much responsibility customers have has changed subtly, he adds. Big data analytics is stunningly easy to consume, once a customer knows how to get their data to the cloud as well as how to describe it using a service in a meaningful way. "Thus customers are able to gain disproportionate amounts of value in the cloud, which should also be used as an acid test for cloud economics."

Hyperscale really is the answer to the age-old IT request to do more with less, says David Odayar, datacentre architecture lead at Comstor Sub-Saharan Africa. It not only reduces the reliance on physical infrastructure by putting the intelligence in the software, but it also facilitates speed of delivery, not only of systems, but of applications too.

You don't move into digital transformation, you architect it.

Sven Blom, Teraco

According to Julian Thomas, principal consultant at PBT Group, big data analytics and hyperscale unlock endless possibilities for the African business environment in sectors, for example, like construction, where workload scale and durations are unpredictable. Now, businesses can scale up or down quickly. These innovations also reduce the burden of having to invest in expensive infrastructure.

The changing landscape

The cloud provides an environment where resources can be easily provisioned, scaled up, managed and retired, if no longer required. This is great, except that IT and business executives were approaching their investment committees and shareholders with business cases for building datacentres and developing internal technology engineering capacity only about 10 to 15 years ago, says Zakes Socikwa, cloud, big data and analytics lead at Oracle. And, in some cases, these investments are yet to deliver returns. Today, these same people are having similar conversations with key stakeholders about developing a business case around cloud. This can mean abandoning planned projects and having to deal with costly half-empty private datacentres. But this is the reality modern businesses face, he says. This is especially true as several big players express interest in opening datacentres in South Africa. Microsoft already launched its facilities in Johannesburg and Cape Town in March.

Adoptions hurdles

For PBT Group's Julian Thomas, the slow pace of adoption locally, with many only beginning their journey into the cloud now, it's important to note that moving core operations and systems into a new environment is not something that can happen overnight. He cites the following as potential adoption hurdles to watch:

  • Technical challenges - Depending on the company size, it's important to phase in the transition so that you allow users to gain a 'feel' for the new environment. This also reduces risks.
  • Existing contracts - Often, businesses embark on their cloud datacentre migration initiatives, but come up short because they fail to read the fine print on their existing contracts, which prevents them from moving without incurring major penalties.
  • Human resources - HR must be included in cloud conversations. In addition to this, organisations should not forget that people hold unique skills that can be used to automate other IT functions with great benefit to the business.
  • Security - Always an area of concern, clear policies around security and data regulation must be considered and developer as needed.

Socikwa describes interest from the likes of AWS, Microsoft and Huawei as a good indicator of growing demand. However, the adoption of cloud by South African businesses is still very low, he continues. What these new facilities do is make it easier for businesses to comply with data residency issues. South African businesses are subject to various data regulations, such as the Protection of Personal Information Act (PoPI) and the EU's General Data Protection Regulation (GDPR). Failing to adhere to these regulations can result in hefty penalties or open the business up to various risks.

Mikey Molfessis, cyber security expert at Mimecast South Africa, is also cautious. Local businesses must remain mindful of the security implications of relying exclusively on a single cloud service provider. "You're not only putting all your eggs in one basket, you're putting them in the same basket as everyone else." This is a hugely attractive target for cyber criminals because they know they only have one lock to pick to gain access. They also know that the potential payoff is massive, he notes.

Molfessis advises that businesses develop a cyber resilience strategy much like they would have done when their systems were on-premise. The fact that data is now in the cloud doesn't mitigate the need for businesses themselves to have a plan in place should something go wrong.

And we also can't ignore how local hosting partners will be affected by the launch of these datacentres, notes Blank. Without a key differentiator, they won't be able to compete with the large 'gorillas' in the room that can out-scale and out-innovate them.

A driver of change

Gary Allemann, MD at Master Data Management, has witnessed big data analytics becoming increasingly mainstream and describes it as a big driver of change. As organisations look to find new ways to leverage data, something like on-demand scaling in the cloud becomes especially important. Allemann advises organisations to ask questions around data security and they should be mindful of the costs associated with cloud. Having the bandwidth to sustainably, cost-effectively and timeously move data between internal and public cloud datacentres is also a must.

From a skills perspective, Hoggarth believes the future of big data lies in the hands of the data scientist. Data scientists are disrupting many traditional job roles, he says. We're seeing code playing a massive role in correlating live feeds into dashboards and reporting tools, with an increasingly strong focus on auto updates, the use of AI, cognitive analytics and machine learning to drive value and enable real-time decision-making.

Digital dinosaurs vs digital predators

Forrester Research describes the combination of big data and public cloud as an extinction-level event for digital dinosaurs. The digital predators that get there first will exploit innovation in the public cloud and become more customer-obsessed. The digital dinosaurs will react too late, scramble to win back customers and eventually die off.

But modern organisations should know that big data analytics in the cloud is irrelevant by definition, says Hoggarth. This is because big data intrinsically encompasses data not only in the cloud, but also on devices, on-premise and in transit - big data includes every sort of data. Yes, cloud is a powerful enabler, but it's not the be-all and end-all, he asserts. What organisations should be focusing on is connecting all the dots and recognising that they're now operating in an environment where data is continually moving across multiple technologies, siloes, formats, locations and vendor environments.

Sven Blom, head of sales at Teraco, agrees, and acknowledges that big data analytics and hyperscale are key enablers of digital transformation. He believes the future is undoubtedly hybrid. However, he adds, none of the benefits will be realised if your infrastructure and network aren't scaled and built with digital transformation in mind. Locally, there are still far too many businesses that need to get the network and infrastructure right before even considering what big data analytics can do for their businesses, he continues. The cloud is only as good as the network it's running on, so it's critical to address latency and sovereignty issues before even getting started on the cloud journey.

"You don't move into digital transformation, you architect it."

Figuring out how to do so, and finding the right tools and solutions to help you along the way, can be overwhelming, notes Andrew Cruise, MD at Routed. This could be considered both an advantage and disadvantage of digital innovation because the sheer number of technologies and services available means that modern businesses really are spoilt for choice, but it may also delay their decision as they agonise over finding the right fit.

You're not only putting all your eggs in one basket; you're putting them in the same basket as everyone else.

Mikey Molfessis, Mimecast South Africa

In the end, it's up to each organisation to step up, build a robust architecture that can support required capabilities, implement this strategy and take full advantage of what the technology can offer, says Oracle's Socikwa.

Organisations shouldn't allow themselves to be held back by technology, cautions Hoggarth. They should be driven by the business needs, recognise why they're doing it in the first place and focus on the value propositions/constraints. Today, businesses have to keep in mind that yesterday's principles won't make the grade. It's imperative to plan with current trends and innovations in mind, he says. These changes are happening around and to your company. Avoid them, at your peril.

Share