Whether we like it or not, the concept of hybrid cloud is getting its share of attention from the IT community. In fact, most organizations seem to be either thinking about—or actively trying to move to—a hybrid cloud model. We see this often in our meetings with customers. We also see many IT shops struggling to identify and master the basics. There are core elements that need to be understood about the hybrid cloud in order to successfully transform the organization.
We field a number of hybrid cloud questions. Customers want help understanding hybrid cloud. They ask how they can use it to regain control of their IT infrastructure. They also want to know how they can use it to run workloads for short durations and avoid CAPEX expenditures.
If you’re wondering about hybrid cloud, this article can help. In it, I offer five fundamental areas of expertise your IT team should learn and master now in order to better prepare for a hybrid cloud transformation.
BUT, FIRST, SOME “CLOUD COMPUTING 101”
Before we tackle my list of five, I’d like to get a few cloud computing basics out of the way I have heard too many variations on what people believe cloud is. So, I’d like to describe how virtualization fits into cloud environments, as well as the connection between cloud and IaaS. I’ll also discuss one way to think about what a hybrid cloud really is.
1. Virtualized Infrastructure – Do you run virtual machines in your data center, based on VMware (or, maybe Red Hat KVM or Microsoft Hyper-V/System Center)? If so, you already have a virtualized infrastructure. You also already have the foundation for a cloud-like architecture based on Infrastructure as a Service (IaaS). Despite all the hype these days around Software-Defined Data Centers (SDDC), it might surprise you to know we’ve already been running ‘software-defined’ virtual servers for the past 14 years in our data centers! When your virtualized infrastructure can be trusted to run efficiently in your data center, you already have the start of an IaaS platform.
2. IaaS – Most enterprise-level discussions about cloud computing (private cloud, public cloud, or hybrid cloud) tend to center around a specific type of cloud computing: Infrastructure as a Service (IaaS). IaaS can be found in internal private cloud environments, in public cloud environments or in hybrid clouds. VMware vCloud Air and Amazon Web Services (AWS) are good examples of public cloud providers who offer IaaS. They offer various infrastructure resources, such as virtual compute or infrastructure, as an on-demand, pay-as-you-go service. In contrast to public IaaS, private cloud IaaS might involve a company’s internal IT department that offers on-demand services to end users (or key lines of business). Such services may also be comprised of select compute, network and storage resources, often bundled together. Granted, many private IaaS implementations might not go so far as to charge individual departments or end users for the use of their services, like their public cloud counterparts do. But, an IaaS model’s measurement/metering capabilities can still provide IT teams with a deeper understanding of the cost involved in delivering services—whether those services ultimately come from their own private cloud environment or a public cloud provider.
3. Hybrid Cloud – Here’s one way to think of a hybrid cloud. Take your existing virtual infrastructure, then add efficiency, consistency, cost control, and understanding. These characteristics are the makings of IaaS. All these IaaS characteristics must be present for this next, hybrid cloud generation of IT operations.
Now, along with your own IT organization offering IaaS to its customers, imagine adding resources from a public cloud service that also provides its IaaS resources to you. You now have the option to understand which cloud provider’s IaaS resources to choose from (i.e., your private cloud’s resources or someone else’s public cloud resources) when it comes to deploying and running specific workloads. The idea of running some workloads in your private cloud while also running other workloads in one or more public clouds is the essence of hybrid cloud.
THE FIVE AREAS TO MASTER FOR YOUR FUTURE HYBRID CLOUD
If you’re serious about making a successful move to hybrid cloud, now’s the time to make progress in each of these five areas.
1. Validated architectures and best practices.
This seems like an obvious area to start with, but it’s more critical than ever to the emerging models of IaaS and the enablement of private and/or hybrid cloud environments. You should start by assessing how your current virtual infrastructure and overall data center operations are doing in regards to architectures and best practices. Getting an outside perspective on where you stand in these areas can be valuable as well. Many companies we see tend to follow nonstandard, inconsistent processes for everything from setting up an operating system to tagging a VLAN. Starting with a validated foundation, that follows industry best practices, will make it easier to automate key processes you need later for a successful IaaS and hybrid cloud deployment.
2. Accurate IT cost modeling.
As a future cloud service broker, your IT team will need to accurately assess and evaluate the performance and cost of cloud services provided to your users. This includes costing information, such as the cost per VM and the cost per workload associated with certain infrastructure services or applications. This will help you get a better handle on the current cost involved in offering cloud services from within your IT data center. It will also help you compare external cloud providers and develop more specific SLA requirements. This is an area you should be refining now with your current virtual operations.
3. Efficient virtual operations.
This area relates to the last, cost-modeling area. Many organizations are naturally evolving from virtual infrastructures to IaaS and private cloud, with an eye toward hybrid cloud down the road. As you assess the costs of running your current operations, including costs per VM and costs per workload, you may identify obvious areas of improvement. There may be other areas of your virtual infrastructure that require an expert to analyze potential costs per workload and advise where to make certain workload-specific performance or cost improvements. Believe it or not, focusing on these storage types of efficiencies can mean the difference between an internally-supported virtual application that costs $100 to run vs. an inefficient application operating in a public cloud for the high cost of $5,000!
4. Lifecycle management.
One benefit of IaaS is the ability to do more with less. To achieve that benefit, however, also requires better management of existing physical and virtual resources. (How many of us have physical servers or virtual machines that we are just not sure who owns them?) This area includes retiring those resources that no longer provide any level of value. It also involves smarter resource acquisition. (If we purchased cars like IT organizations have routinely purchased hardware, we’d end up with 15 cars in the driveway!)Effective life cycle management assesses application lifecycles. It can involve retiring less-critical applications after 5 years. It may also involve other discussions, such as the need to move older applications from more costly, 1st-tier storage to a more affordable, 3rd-tier storage.
5. Cloud automation.
IaaS and the cloud represent a brave new world for IT teams accustomed to working with other teams to manually fulfill project requests for new storage, new VMs, or new environments. We’ve been hearing about cloud automation and orchestration for a few years. We feel that today’s vendor tools associated with automation and orchestration have since reached a maturity level that makes them easier to implement and deploy with minimal scripting. Now’s the time for such tools—often developed by virtualization vendors—to be tried in early IaaS and private cloud pilots. These are “the glue” that will ultimately tie together the management and oversight of private and public cloud services under a single, hybrid cloud architecture. These will also be the central theme for IT’s hybrid cloud transition to the role of broker of cloud services.
Does mastering these five areas mean you’ll never face a snag in your move to hybrid cloud? Of course not. Adding hybrid cloud as an additional IT resource is a new way of doing things. It’s a new architecture, complete with new processes and new mindsets needed for all involved—from the CIO to application support, network administrators and end users.
What mastering these areas will do, however, is accelerate your cloud learning curve and shorten the time it takes you to get there.
