It’s not just about choosing the right cloud for a business’ current needs, it’s about having the right data management in place to ensure business ambitions don’t outgrow cloud architecture, says Huw Owen, Head of EMEA & APJ at Couchbase.
From brave new approach, to industry buzzword, to the foundation of many critical services – cloud has come a long way in a relatively short time. Even looking back just ten years, early cloud adopters were mostly concerned about two things: simplicity and cost.
Those pioneering organisations typically deployed a small-scale integrated stack from one cloud provider and used it to host relatively simple applications, such as back-up or chat apps. And it worked brilliantly. Organisations were suddenly able to create tools and services that simply weren’t possible before, or at least, couldn’t be delivered at the necessary scale and cost.
When it comes to application performance, most users today expect applications to be extremely responsive, often with sub-millisecond response times
Fast forward to today and organisations are now under enormous pressure to create highly complex, real-time responsive applications. Indeed, in a recent digital transformation study, 85% of IT leaders said digital disruption in their industry had accelerated over the past year. In response, these businesses are looking to revolutionise their digital offerings, with almost 90% of IT leaders admitting they must adapt digital services over the next 12 months or accept they will be less relevant.
The data demands from these new services are huge; from TV services that need to deal with massive spikes in demand when the latest episode launches, through to retailers that need to update in real-time inventories that run into the tens of millions of products.
While the ambitions for more complex applications and uses of data continue to build, many organisations now find that their legacy infrastructure – whether cloud or on-premises – is holding them back, particularly when it comes to how they can use their data.
Three modern data needs
Data is the lifeblood of digital business. It drives competitive advantage and is necessary at almost every point of engagement, from tracking goods in production lines and streamlining supply chains, to understanding stock levels in a store or providing a more personalised service to customers.
Data underpins all applications in today’s massively interactive businesses. The problem is, relying on a single cloud can make it difficult for modern digital businesses to meet their three key needs of flexibility, risk mitigation, and innovation.
While any given cloud provider may be a good choice one year, there is no guarantee that it will meet the business’s cost and resource needs in perpetuity. Not all clouds are created equal; it’s quite common for businesses to want to move certain data or workloads between different clouds to optimise costs.
Clearly, this becomes a challenge when data gets ‘locked’ in one cloud because the provider has intentionally made it hard for users to migrate to another. When companies need to respond to requirements such as spikes in demand, changes to applications, or just long-term growth, they want to be confident that it’s easy to move their data to the cloud or on-premises system that best meets their needs.
When it comes to application performance, most users today expect applications to be extremely responsive, often with sub-millisecond response times. Even small reductions in performance can leave customers frustrated. If a cloud provider was to suffer a major outage, it could lead to a service not being able to work at all and this could have a very public and severe impact on the organisation’s reputation. Although the likelihood of such an outage is small, the threat alone means many businesses want to be able to divert workloads from one cloud provider to another at very short notice.
Organisations are constantly looking for ways to do more with their data; to create or modify applications and services to react to customer needs, new regulations, and their competition. For example, more businesses are looking to push data closer to the edge, so mobile applications and IoT devices can deliver a real-time and ‘offline-first’ experience. A centralised cloud architecture typically doesn’t meet the need to store and process data in this way.
Companies are finding they need to not only easily move data from one cloud to another, but also move data to and from an on-premises deployment to a cloud, and increasingly, from a cloud or on-premises deployment out to any mobile or edge devices.
Adapt or fail – building a cloud strategy that delivers
Growing numbers of businesses that adopted cloud the best part of a decade ago are now adapting their cloud architecture to take a cloud-neutral or multi-cloud strategy; underpinning applications with database technology that works across every cloud provider, and also choosing the best cloud for each business need – which often results in multiple private and public clouds being used.
Let’s say, for example, that the cost of an organisation’s public cloud starts to spiral out of control. With a cloud-neutral strategy in place, the organisation could build out their own data centres and run their applications in their own private cloud, with no extra effort.
The same is also true for a new data regulation. Multi-cloud makes it possible for a company to move any affected data and workload into a secure on-premises environment, without the headache of cloud lock-in.
Businesses are faced with the complex reality of architecting their cloud to meet the changing needs of the business
But while it delivers on some of the modern data needs, multi-cloud is not without its challenges. The obvious problem with having a more diverse environment is how to stitch all these parts together so your data isn’t siloed. APIs can help to provide access to data across platforms, but these often come at the cost of performance and complexity, something that innovative digital companies simply can’t afford.
One way that organisations are adapting their multi-cloud set up to deal with these challenges is by building an independent data layer to act as a distribution centre – delivering data to applications whenever and wherever they need it. The data layer works alongside any mix of public or private clouds. It no longer matters where specific data resides – if a business wanted to maintain a user’s profile across multiple channels, they can.
For example, a retail customer might start browsing via a mobile application that is hosted on public cloud infrastructure, add extra items and checkout via the retailer’s website based on a private cloud, and finally use in-store digital displays, running off an in-store server, to complete their click-and-collect order. The data needs to be consistent across all of this infrastructure to ensure that the customer experience is seamless.
Cloud is a journey, not a destination
Cloud architecture, like many aspects of business, is a journey not a destination – constant refinement and adaptation is key to avoid stagnation and failure. Businesses are faced with the complex reality of architecting their cloud to meet the changing needs of the business. Where data is concerned, they must be able to abstract data from their clouds and deliver it wherever it is needed, in as close to real-time as possible.
It’s not only about choosing the right cloud for the company’s current needs, but also about having the right data management in place to ensure that business ambitions don’t outgrow cloud architecture.