The evolution of storage
Pay-per-usage is becoming the norm.
Interviewees: Rohan de Beer, Jan Sipsma
Storage is having to evolve as businesses are finding it increasingly difficult to calculate what their storage requirements will be in future, and data grows at an exponential rate. Organisations worldwide are trying to save as much data as possible, so that it can be used to drive business value. A new business model that's coming to the fore enables businesses with huge data growth and a cloud strategy to access a cloud environment on a pay-as-you-use basis. iSanity COO Rohan de Beer says: "Simply put, the business pays for its usage and the amount can vary monthly."
According to De Beer, the trend towards Storage as a Service on a 'Pay-as-you-Use' model is being driven by big data, IOT and budgets that aren't growing at the same rate. He says: "Traditional storage based on known architecture that's been tried and tested has been more than adequate until now. However, this legacy approach to storage has had to evolve into a more agile technology that's built around machine learning, as the infrastructure needs to be intelligent enough to learn what the customer uses, to understand the environment and adapt in real time to benefit the user."
Businesses today need to be dynamic and quick to market, they can't allow infrastructure to hold them back, and that means an infrastructure environment that can support the Fourth Industrial Revolution. It needs to be agile and traditional systems just can't cater for that. De Beer says: "Legacy methods aren't able to take a business into an environment in which decisions need to be made on the fly based on analytics and in which artificial intelligence is all-pervasive. Both of which are reliant on the ability to store and process big data."
Jan Sipsma, Solution Architect, Infinidat, says: "Big data is often superfluous data that customers aren't necessarily using to help them build their future businesses. We assist our customers to use their valuable data to generate information that can help them make better business decisions."
Data analysis is particularly lacking at SME level, continues Sipsma. "People don't use data to its full extent to help grow their business. We empower businesses to extract more information out of data so that they, in turn, can better service their customers."
He cites the example of an agricultural cooperative that has access to its many customers' agricultural data. "By enabling the co-op to analyse all the data that it has at its disposal, it can in turn provide valuable information to its customers around trends that emerge that might impact on them in the future." This represents a new generation of highly optimised storage that enables organisations across all industries to achieve petabyte scale, at a fraction of the total cost associated with legacy storage.
De Beer says: "Before customer can analyse their data, they need to identify redundant, duplicate data or obsolete with no value to the business. All too often businesses simply don't have visibility into their historical information. They need to decide which is current data, what can be archived and what needs to be retained for compliancy purposes. It's all about structuring the data and providing visibility into what data is in the environment. There's simply no point buying more hardware to host their data if that data is not needed."
Once a business has decided which data to store, which data to analyse and which data it accesses daily, it's ready to adopt a pay-as-you-use storage model, says Sipsma. "This is done in one of two ways. It can be provided via a virtual data centre (i.e. in the cloud) and the customer is billed per usage on 'Infrastructure as a Service' and data protection services. The alternative is an enterprise box installed at the customer's premises that comes with a capacity-on-demand usage model. The customer has full access to the entire storage unit but is only invoiced based on usage."
A challenge faced by a lot of enterprises is that they can estimate their future data growth based on statistics that make predictions based on historic trends but, according to Sipsma, this is at best a guess and businesses generally must budget for unforeseen additional capacity 'just in case', regardless of whether they use it but customers still pay for extra capacity from day of implementation.
The evolution of the pay-as-you-grow model means that businesses will always have spare capacity available, but will only pay for it if they use it, he explains. "It also means that the business can access that storage at a moment's notice, which is important in today's fast paced world. From a Capex perspective this is very appealing to CFOs from a budgetary perspective."
This approach gives businesses the flexibility to take on new projects at a moment's notice, without having to wait for additional capacity to become available - or for approval of the budget to purchase it. This is a vital component of being first to market in a highly competitive economy.
De Beer says: "We're anticipating massive data growth over the next five years, with zettabytes of data being generated by 2020. Businesses need to consider the challenges the data centre will face around scalability in the future - and plan accordingly. They need to seek out an architecture that will enable them to embark on a long-term data strategy."
Which is why we're seeing the emergence of trends such as Storage as a Solution, Access as a Service and Architecture as a Service, to ensure that massive amounts of data can be analysed faster. Sipsma says, "Businesses don't want to buy a functional solution per challenge, they want a single, multipurpose device that can give them everything from slow storage as a service to high performance storage as a service and the flexibility to switch between the two."