By Michael Thompson
It’s no secret that virtualization rules the data center—it just recently surpassed 50 percent of all server workloads and is expected to reach 86 percent by 2016. But although virtualized data centers have largely become mainstream, the majority of organizations are still in what might be considered the early phases of virtualization maturity.
So, how can businesses navigate this landscape to fully realize the benefits of advanced virtualization technology? Today, there are four important phases in the virtualization maturity progression: implementation, optimization, automation and advanced automation, which encompasses bleeding-edge enhancements such as software-defined networking and storage.
What follows is a guide to each phase, including best practices that will enable IT organizations to most effectively employ virtualization today and keep pace with its future applications.
Phase 1: Implementation & Baseline Operations
To get up and running, IT organizations should prepare for an effective implementation by establishing core operational technology and procedures related to proper configuration, security, migration, disaster recovery and backup processes. The more mature and stable these capabilities are before virtual machines (VMs) are created and workloads allocated to each, the easier the implementation will be. After initial deployment, virtualized servers must then be closely monitored and maintained to fix inevitable performance and capacity problems as well as to document a baseline of operational efficiency. This type of laser-focused monitoring also helps ensure availability and the prioritization of resources to best meet the needs of SLAs and end users.
All too often, this phase is plagued with errors and oversight, such as improper capacity planning or establishing a virtual environment without the expertise to eventually scale and enhance the technology. These kinds of common implementation missteps can create a host of often expensive problems that ultimately hinder an organization’s ability to move to the next phase.
At this point, most enterprises have gone beyond basic implementation and have established a steady-state virtual environment. Many of these same organizations, however, often never move beyond reactive management techniques and can become permanently stuck in this first phase. The IT organizations that are asking themselves, “How can we do this better?” are those who understand the benefits of moving to proactive management and have progressed or are now progressing to the optimization phase, which can significantly increase agility and performance.
Phase 2: Optimization
After successfully deploying virtualized infrastructure and reaching a baseline level of operations, IT organizations can turn their attention toward optimizing that infrastructure for performance and agility while ensuring the business is getting the most out of its physical hardware and realizing a noticeable competitive advantage with the technology. But the freedom that virtual environments provide also presents unique challenges that administrators must manage to maintain an optimized environment—in addition to requiring detailed planning and attentive management.
Organizations can enhance the performance of a virtual environment in several ways, starting with simply consolidating VMs for significant capital, operational and energy-consumption savings. But for enterprises that require a larger virtual environment, managing sprawl and right-sizing VMs are paramount to optimization.
Virtualization makes the creation of additional virtual machines easy; overprovisioning, however, can lead to a significant drain on the overall performance of the environment. Unchecked VM sprawl also often results in security breaches if old, forgotten VMs are left unpatched. To help stem the problem of VM sprawl, IT organizations should constantly pose the question, “Is this VM still required, and if so, does it have the right level of resources?”
Other best practices for managing sprawl include implementing a formal request and approval process before creating a new VM, limiting which IT admins are allowed to create new machines and establishing a detailed chain of ownership for created VMs. Organizations can also establish a showback system, where reports on resource consumption—and possibly even theoretical costs—can be shared with end users to encourage good behavior and demonstrate that VMs are not free.
Aside from controlling sprawl in a virtual environment, administrators must also plan for any future needs that business may have for additional VMs. As such, capacity planning is also an important part of the optimization phase to make sure enough capacity exists across all host resources, and it requires an organization to have a thorough awareness of historical performance data to make informed recommendations. These optimization processes create order out of the chaos that can often overtake a virtual environment, allowing organizations to easily progress into the third stage.
Phase 3: Automation
For organizations ready to move beyond optimization, the third phase of virtualization maturity is automation. This stage relies on an enterprise having a strong foundation of basic processes, in addition to well-understood and well-documented workflows to enact best practices and policies that enable automation for scalability and agility improvements. These processes should be established in Phases 1 and 2 of the enablement journey.
Truly optimizing a virtualized environment starts with instrumenting the environment with an automated management tool that gathers data, analyzes performance and provides automatic alerts that then form the basis for more-advanced automation and orchestration. These alerts can be self-automated, in which case a user validates an action before “clicking the button,” or fully automated, in which a machine will take action on its own. Organizations that successfully move to automation will create a more flexible and responsive virtualized infrastructure that truly unleashes the benefits of virtualization, including enhanced speed, greater cost savings and simpler end-user servicing.
But determining when and how to begin automation can be difficult. A good approach for first timers is to phase in automation on the basis of risk, value and predictability. In other words, play it safe at the outset. Identify a test environment that is low risk, high value and as predictable as possible to learn from mistakes and create best practices for wider deployment. IT organizations should also pick an automation tool or framework that matches the skills and expertise of the organization’s subject-matter expert.
Phase 4: Automation 2.0—Software-Defined X
Although Phase 3 (automation) is the goal for most organizations on the path to an effective virtualized environment today, the industry has already started looking toward Phase 4, or what is considered the bleeding-edge of virtualization maturity: software-defined networking and storage. These technologies enable businesses to accelerate application deployment and delivery beyond just compute resources to dramatically reduce IT costs through more data-center-wide workflow automation.
Software-defined data-center technologies can be easy to buy, but the hard part is adapting current processes and technology to effectively employ them so they truly make business more effective and efficient. Before contemplating a move to this stage, IT organizations should ensure they and their virtual environments are built on solid foundations of best practices from each of the three previous phases. They should also see a clear competitive advantage to using this technology that justifies the risk and expense.
Virtualization’s benefits are indisputable, delivering cost savings, increased availability and improved agility to meet end users’ needs. But for many businesses, expanding beyond the most basic implementation of the technology proves difficult. Whether an organization is just beginning its virtualization journey or looking to progress to a more advanced stage, following the framework in this guide can empower it to not only maximize the benefits of the virtualization phase its currently in, but keep advancing to the next phase and realizing the added benefits available.