Why a Vertically Integrated Cloud Stack Could Be the Wrong Move

March 16, 2018

When the cloud first entered the enterprise consciousness, the pitch was simple: move your applications to the cloud and enjoy the simplicity of an on-demand, integrated stack without the hassles and headaches.

There is a lot of merit — and allure — to this promise. And, as organizations began experimenting with small, greenfield-type applications, it worked magically.

But as organizations are beginning to adopt the cloud at scale, they are realizing the unspoken trade-offs they were making by moving to a single-cloud, integrated stack solution – particularly when it came to their data.

As enterprise leaders enter this next phase of cloud adoption, a new set of critical success factors are emerging: flexibility, the ability to innovate, and data portability. These organizations are beginning to wonder if a single, integrated cloud stack may have been the wrong move.

The Need for Flexibility, Innovation, and Data Portability

While the early drivers of the cloud movement were all about cost-savings and agility, which are still important, the business motivations for using the cloud have evolved.

Today, organizations need to go beyond merely having applications in the cloud. They must instead use it to create and enable innovations that deliver competitive advantage. This imperative requires more diverse and often more complex and data-intensive applications.

Creating advantage from cloud-based applications, however, requires that organizations seek to innovate across every layer of the stack – as each layer is now integral to today’s dynamic and performance-sensitive applications.

While innovation at each layer will be critical, organizations should pay close attention to a foundational layer that they may have taken for granted: the database.

Data is the lifeblood of any organization and is now one of its chief drivers of competitive advantage. What’s more, the speed and demands of the market are pushing elements of the technology stack closer to the edge in the form of mobile devices, IoT and the like.

In this environment, in which data holds the veritable ‘keys to the kingdom’ and in which it is necessary at every point of engagement — in the data center, in the cloud, and at the edge — the scalability, locality, and portability of that data becomes essential.

The challenge with the single-cloud, integrated cloud stack then becomes apparent: it effectively replicates many of the data challenges of on-premises systems, just in a modern edifice.

Data Abstraction and Multi-Cloud

To overcome the challenges of data lock-in and to create the type of architectural flexibility that the modern enterprise requires, organizations must seek to adopt two new conceptual models: multi-cloud architectures and data abstraction.

The idea that they must adopt a multi-cloud architecture — in which they integrate multiple incarnations of public cloud instances from different providers, along with private cloud instances — is becoming widely accepted as organizations mature their cloud deployments.

There is a growing playing field of both general-purpose and specialty public cloud providers, along with a vast range of private cloud architectures available to enterprise organizations. Each of these approaches has their strengths and weaknesses.

Stitching these multi-cloud environments together, however, becomes difficult when the most common single cloud deployments simply use the integrated data layer. While organizations can and do use APIs to provide access to data across platforms, this often comes at the cost of performance and complexity.

As a result, organizations are turning to data abstraction to help solve this problem in a multi-cloud world. The idea behind data abstraction is to create a data layer that is independent of the public or private cloud stack, but which can deliver data to applications when and where they need it using techniques such as advanced data replication and offline sync.

Abstracting the data from the integrated stack enables organizations to use a mix of public and private clouds in the way that is most effective for them. Beyond this flexibility, data abstraction also overcomes the other major challenge with a multi-cloud approach: complexity.

Technology vendors and industry observers often talk about multi-cloud as if each public cloud provider is the same. In reality, however, each provider takes a very different approach in how they implement and instrument their stack. As a result, enabling transient workloads in a multi-cloud environment is very difficult.

Data abstraction hides this complexity and makes the use of a multi-cloud architecture tenable by providing a consistent data management layer.

The Intellyx Take

There’s a big story playing out inside enterprise organizations today. They are rapidly evolving past the wide-eyed fear and excitement that embodied the early days of digital disruption and the resulting call for transformation.

They are now settling in to deal with the complex reality that they are facing: they will require every resource at their disposal to make things work. Whether you call it hybrid cloud, hybrid IT or multi-cloud, this is their new reality.

The need to leverage their data ubiquitously and rapidly across these multi-cloud environments is, therefore, driving enterprises to seek out modern data platforms, such as Couchbase, that will enable them to abstract data and deliver it whenever and wherever their business needs demand.

Creating this data abstraction layer and the capability to leverage data across a multi-cloud architecture will be critical to enabling a consistent user experience and the organizational flexibility needed as organizations drive their digital transformation efforts.