Article (4 minute read)

The Difficulties in Weaving Together the Loose Threads of Multicloud

Multicloud is the future of computing. For most organizations, it’s the present too.

Jeff DeVerter / Rackspace

Multicloud is the future of computing. For most organizations, it’s the present too. When asked how many clouds are in use, most organizations would answer only one. Yet, in reality, statistics indicate that across most organizations, there are an average of five clouds in use. Even if you’re office is entirely Microsoft-based, you’d still be considered a multicloud company as Azure, Office 365 and Dynamics workloads live on different cloud management systems.

For context, let’s define multicloud as embodying a variety of cloud types (public, private and on-premises) from a variety of cloud providers (e.g. AWS, Google Cloud Platform and Microsoft Azure) and/or SaaS providers (e.g. Salesforce, SAP, Oracle) in the same environment. These loose threads can lay the foundation for an integrated fabric that drives innovation. Conversely, without proper strategy and execution, they can become frayed, inefficient threads that inhibit innovation.

As an organization matures, it becomes nearly impossible to get everything from one vendor or run it all in the data center. This is where a multicloud consumption model emerges. Organizations looking to leverage AI or machine learning for predictive analytics need a cloud with high data processing capabilities. On the other hand, an organization looking to capitalize on containers might choose a cloud with more evolved container deployment features. You don’t have to be constrained to a single cloud or be concerned about vendor lock in. You can place workloads where they run best.

Where the ends fray  

If moving to one cloud is good, then two clouds must be better, right? You can save your organization from the evils of vendor lock-in, expand the technology landscape for your technical staff and provide technology choices to business units with a favorite cloud provider. These all sound like good and noble reasons to adopt multiple clouds into your computing landscape. 

But I have a warning for you: 1 + 1 does not equal 2. The amount of effort it took to adopt your first cloud multiplies when adding a second or third cloud to your corporate IT landscape. Due to the effort involved and the reach of its impact on operations, your decision to pursue multicloud should be based on an honest look at several hard-to-manage criteria. Let’s address a few of those:

Governance 

Cloud governance is the who, what, where, why and when of your clouds. Governance helps you define, permit and enforce who can do what in the cloud, under what circumstances and when they are allowed to do it. This is a relatively easy thing to do in traditional infrastructure as there isn’t much variation in what you can do with it. But in a cloud environment, the options are almost limitless. Different clouds follow different governance protocols. The challenge lies in tying all of those cloud governance requirements and capabilities together to enforce an enterprise-wide governance solution. 

Inventory 

Keeping track of cloud assets is difficult to do in a single cloud. In a multicloud environment, it’s even more complex. Yet, it’s incredibly important for a company to have a solid grasp of this for a multitude of reasons including financial, security, regulatory and disaster recovery. Though each cloud provider has native tooling to accomplish this task, in a multicloud environment, it’s incumbent on you to maintain this data. With this data, you can facilitate consistent cross-cloud tagging, enable unified billing and achieve more effective chargeback/showback reporting.

Event logging

Having a database of all active assets running across all of your providers is only the first part of the task.  Next you need to understand exactly who is accessing them and what is running on these technology assets.   Having a method capturing this data across all cloud facilitates the ability to create dashboards and automation that triggers on a failed event, missing software, configuration shift, and of course – security and compliance requirements.

To meet security and compliance regulations, you must log events and, through careful analysis, remediate and prevent further events. The recent increase in remote working increases the need to track endpoint devices and automate disaster recovery. By its very nature, a multicloud environment generates a broader attack surface. And you need to see the full picture. To do so, you’ll need the capability to collect, connect and report across all of your cloud environments.

Despite these challenges, I’m not recommending that companies avoid a multicloud strategy. In fact, as stated above, I believe that multicloud is the future — and present — of all cloud computing. However, I want to stress that adding a new cloud to the mix needs to be a thoughtful and purposeful activity tied to serving business needs. Like any technology consideration, it should be applied to a specific business challenge or a business opportunity. Be cautious not to let the cloud pendulum swing too far in either direction — you don’t want to be locked into a hyper-controlled, cloud experience or, even worse, drowning in an uncontrolled, multicloud soup. Be prepared to build the tools that you’ll need to manage your multicloud environment, or seek out a provider who has already solved the multicloud management dilemma.

Join the Conversation: Find Solve on Twitter and LinkedIn, or follow along via RSS.

About the Author

CTO, Products and ServicesJeff DeVerter

Jeff has 25 years of experience in IT and technology, and has worked at Rackspace for over 10 years. Jeff is a proven strategic leader who has helped companies like American Express, Ralph Lauren, and Thompson Reuters create and execute against...

Read More


Solve Strategy Series

Register for one or all of these global events, featuring industry influencers, experts, technologists and leaders

Register Now