Embracing the Hybrid Cloud Challenge
January 29, 2016 No CommentsFeatured article by Kent Christensen, Datalink’s cloud and virtualization practice director
The concept of hybrid cloud has been around since organizations first began to embrace cloud computing, but most IT teams are still struggling with putting the concept into action. The definition of hybrid cloud is something of a moving target. So is the technology currently used to build and manage them. In this article, we’ll briefly address these issues and provide some guidelines for beginning your hybrid cloud journey.
The Promise of Hybrid Cloud
If you asked one of your peers what a hybrid cloud is, you might get answers like “a company’s use of both internal (private cloud) and external (public cloud) resources.”
While this answer is accurate to a point, the promise of hybrid cloud goes far beyond the use of popular public cloud SaaS solutions like Salesforce.com in conjunction with an organization’s own internal private cloud infrastructure.
The real value of hybrid cloud lies in the simultaneous use of multiple, heterogenous, internal/external cloud resources to deliver some form of underlying Infrastructure as a Service (IaaS). This involves provisioning and common management of an underlying layer of largely virtual compute, storage or network resources that can be simultaneously sourced from your IT organization’s own private cloud environment as well as from one or more external public cloud providers.
Gartner’s hybrid cloud computing definition touches on this wider promise. It states, “Hybrid cloud computing refers to policy-based and coordinated service provisioning, use and management across a mixture of internal and external cloud services.”
Shopping for Hybrid
Some key things to aim for in your future hybrid cloud might include:
– The use of policies to guide and control hybrid cloud use and access. This includes the use of common network policies and common security policies across disparate cloud applications, regardless of whether elements of the applications are based in a public or private cloud.
– Unified management and oversight of hybrid cloud resources. This might involve some form of central or overriding coordination and orchestration that aids in the effective management and global provisioning of combined cloud resources. Gaining central visibility—even single-pane-of-glass management— across private and public cloud resources would also be important.
– Workload or data portability between clouds. Here, the ideal to aim for is secure, bidirectional mobility of application workloads or data across heterogeneous clouds. This includes the ability to globally provision, manage and transfer workloads, as needed, to whichever cloud is the best fit at that particular time.
While technology continues to emerge in this area, I’ll mention three commercial hybrid cloud options available today: Two are in the hypervisor arena and one is in the networking space. The choice of solutions has as much to do with your organization’s preferences as it does with the robustness of a specific vendor’s hybrid cloud offerings.
On the hypervisor side, you have VMware’s vCloud Air ecosystem which builds on the company’s leading virtual server leadership in the enterprise. Microsoft also offers hybrid cloud functionality utilizing Azure Pack and Hyper-V in private cloud with VPN extensions into Microsoft Azure public cloud.
On the networking side of hybrid cloud, you have Cisco and its strong installation base in private cloud environments. Here, for extension of existing network/security policies to external clouds, Cisco recommends the use of its Cisco Intercloud Fabric based on its own ecosystem of providers.
Some Hybrid Cloud Rules to Live By
Within Datalink’s Cloud Complete advisory function, we routinely help organizations “kick the tires” regarding current private, public and hybrid cloud technologies. We also educate them regarding what’s possible now and what’s still coming down the road.
Based on these experiences, I can offer a few other goals to keep in mind as you move forward with your own hybrid cloud:
1. Decide what you want hybrid cloud to achieve for your organization. For example: Do you want to extend your security/network policies to one or more external cloud vendors? Do you want to be able to move workloads from one external cloud to another? How much do you want to leverage your current internal architecture (i.e., porting current virtual machines, etc., between private, public, or public to public clouds)? This will influence which ecosystem you choose.
2. Identify which application workloads are most suited for private, public or hybrid environments. If it’s a hybrid environment, determine the right cloud mix for your applications. Today, you may have to manage hybrid clouds more manually than you might like as you wait for further maturity of automation, orchestration and hybrid cloud ecosystems.
3. Decide to become a broker of cloud services. This means you decide which workloads or data goes to which clouds. You build and/or extend a security and network policy to whichever internal or external cloud resources are required.
4. Consider whether or not you want to maintain choice in your current and future cloud options. For example, if you currently use VMware as your hypervisor but think you might want to try OpenStack or Microsoft Hyper-V later, you may want to consider other hybrid cloud options besides VMware.
Ultimately, deploying a successful hybrid cloud will come down to your organization’s own goals as well as which hybrid cloud ecosystem you rely on to help you achieve them.
About the author: Kent Christensen is Datalink’s cloud and virtualization practice director. He routinely consults with customers regarding strategies and technologies associated with storage consolidation, virtualization and cloud computing.