A Four-Stage Journey to Cloud Computing

Grazed from Sys Con Media.  Author: Jason Cowie.

For almost a decade now, organizations of all sizes have been leveraging server virtualization, but few have fully gained the flexibility and efficiencies it promised. The emergence of cloud computing and the promise of delivering on-demand resources has introduced new challenges and opportunities. A company's level of virtualization adoption maturity correlates to its readiness to deliver an on-demand, real-time infrastructure through a private cloud. An effective virtualization and cloud management solution can ease the journey toward that end...

Virtualization Evolution Stage One: Acclimation
Virtualization evolution stages categorize an organization's readiness to move to the private cloud. The natural path of virtualization evolution moves from the tactical use of virtualization all the way to the optimized use of private cloud services within the enterprise.

This journey from the most basic virtualization deployment to true private cloud use can be viewed in four phases of maturity. The first of these is acclimation. In this early stage, companies are primarily driven by server consolidation and disaster recovery projects. They retrofit their legacy IT processes to manage virtual assets, and they have yet to adopt specialized management capabilities for the virtual data center. At the point of acclimation, data centers still rely on manual, non-standardized provisioning of virtual machines.

In this phase, the journey to the cloud typically travels traditional paths that include legacy processes and non-specialized tools. Usual readiness metrics and indicators that you are in this phase of your journey include:

  • Admin to virtual machine ratios: 1:50 - 1:100
  • Infrastructure density: four VMs per core
  • Provisioning cycle: 15 to 30 days
  • Resource wastage: 35 percent to 50 percent

In this phase of maturity, there are typically four pain points you will encounter and will need to control before you move to the second phase of maturity. The first is the invisible side of virtualization. What insight do you have into your virtual infrastructure? Do you know where your virtual machines are? Without effective inventory management and a means of having complete visibility into your inventory, it's impossible to maintain control or grow your server virtualization environment safely. It is important to manage inventory and gain visibility into your virtual infrastructure.

The second pain point involves tracking. Perhaps you've implemented virtualization, but you don't know where your assets are. With ever-changing virtual assets, ensuring that you have complete visibility into how they are changing over time will give you much more control over your virtual infrastructure. Moving beyond this challenge requires tracking of VM ownership and expiration dates, which can be achieved with a management solution that tracks assets from end to end.

Of course, not all virtual assets are created equally, nor should they be treated as such. The larger your virtual environment becomes, the more complex it is. You need a way to track your ever-changing environment as you scale up and out. Not all of your virtual assets are being used in the same way, so they don't need to be treated in the same way. It's important to create custom attributes to group, view and manage virtualization objects based on business and IT policies.

Finally, virtual infrastructure reporting should deliver transparency. One thing to keep in mind for a successful journey to the private cloud is to ensure that your organization is informed every step of the way. Reporting and notifications can be used to prevent virtual infrastructure problems from occurring, and to provide stakeholders with their own view into the virtual cloud. Customizable reports should help you disseminate information easily to organization stakeholders.

Virtualization Evolution Stage Two: Consolidation
At this point in the journey, organizations are focused on increasing the percentage of virtualized servers and improving consolidation ratios. They are constructing plans for virtualizing mission-critical workloads. The consolidation phase also includes the adoption of specialized management capabilities and processes for the virtual data center. During this time, the data center moves toward a virtual-first policy.

At this phase of maturity, the following indicators of virtualization maturity include:

  • Admin to VM ratios: 1:100 - 1:250
  • Infrastructure density: six VMs per core
  • Provisioning cycle: five to 15 days
  • Resource wastage: 25 percent to 35 percent

One of the top causes of performance degradation in virtual data centers is unplanned or non-approved configuration changes that impact service performance. Change management and auditing are important to the process of controlling the management of your virtual environment. Organizations in the consolidation period should be identifying, monitoring and analyzing configuration changes in their virtual environments and leveraging policy-based architecture to automatically take action to minimize the impact of these changes.

It is also important to monitor performance during consolidation. As more IT services are required, more service requests are created, which means there are more virtual machines and more VM owners, resulting in more configuration management challenges. To keep up with the demands of the business, you need to be able to capture, detect and track actual configuration states against desired configuration states.

At this point, many companies struggle with virtual sprawl. That's a condition characterized by unused capacity on some VMs and capacity shortages on others, resulting in performance issues, unhappy VM owners and excessive and unnecessary costs. Now is when the ability to automatically manage the entire lifecycle of virtual machines from end to end becomes essential.

Finally, moving past the consolidation stage requires companies to maximize their rations. Virtual data center management teams are often thrust into a position where they need to forecast capacity requirements without adequate technology and information to make sound decisions. Capacity management and waste reclamation is an important requirement for any virtualization management solution. Maximizing your resources involves understanding how much system, cluster and host capacity you have remaining and when your available resources will run out.

Virtualization Evolution Stage Three: Transformational
Organizations in the transformational stage are leveraging virtualization as a strategic business enabler. Their virtual data center management capabilities have matured, and they are using business and IT policies to drive automation. At this stage in the journey, IT is ready to adopt basic chargeback (showback) capabilities. We often see standardized processes in place at this point, supporting the provisioning of virtual machines and IT services.

The following indicators are common at this halfway mark on the journey to the private cloud:

  • Admin to VM ratios: 1:250 - 1:500
  • Infrastructure density: eight VMs per core
  • Provisioning cycle: one day to 10 days
  • Resource wastage: 15 percent to 25 percent

This is the moment when having adopted a broad set of management capabilities becomes most critical. It will enable you to see all the various dimensions of the infrastructure from a single pane of glass, which provides the foundation for automating routine tasks, implementing and standardizing repeatable processes, removing roadblocks and inconsistencies, understanding growth patterns and their impact on the business, and prioritizing next steps.

Often, automation gets the highest priority, since manual processes are not only time-consuming, but also error prone and inefficient. Furthermore, manual efforts won't allow you to scale up as your virtualization infrastructure continues to grow. The only way to get to any form of private cloud is through the use of standard practices, policies and automation. Other immediate priorities will likely include request management and empowering select stakeholders with VM management and reporting privileges.

Virtualization Evolution Stage Four: Optimized Private Cloud Delivery
A modest percentage of data centers have reached this endpoint in the virtualization journey, but most have set their sights here. Optimized private cloud delivery is characterized by the delivery of IT services in a consumption-based model that easily scales up and scales down. The service catalogs in such environments deliver services on demand. IT teams have reached mature chargeback and costing models that monitor and curb consumption. Finally, the private cloud provides a fully automated and integrated IT service management scenario.

Such an environment possesses the following characteristics:

  • Admin to VM ratios: 1:500+
  • Infrastructure density: more than 12 VMs per core
  • Provisioning cycle: five minutes to one day
  • Resource wastage: less than 15 percent

Reaching this final stop on the journey to the cloud requires a clear answer to a difficult question: How will you pay for the journey? As virtual environments grow, organizations can support chargeback and other IT costing initiatives by monitoring virtual assets, the resources they use over time and the costs they incur. Even organizations that do not have a formal chargeback program still need to understand the relationship between increased demand, capacity and costs in order to more effectively manage virtual infrastructure growth.

With questions of costs addressed, private cloud organizations must develop interdependent teams that support the alignment of business and IT objectives and the integration of processes and technology. They must fully embrace the IT-as-a-service model and adopt end-to-end lifecycle management and automation. It's a lot to achieve. But many companies are progressing toward these goals at a rapid rate.

Automating the Private Cloud: a Case Study
Aston University, an institution in England, began its virtualization journey in 2004 with a server consolidation initiative that focused on 10 finance applications. Growing organically, this initial small environment became the start of a full-fledged private cloud initiative. Aston moved its first customers onto the private cloud in 2009, billing them through a manual chargeback system. This service chargeback allowed the university to fund the gradual expansion of its infrastructure.

The university established a cloud-service-first directive in 2010 with nearly all new services going automatically to the organization's cloud environment. As Aston's private cloud environment grew, the IT server team started to experience problems with capacity and with the manual chargeback system. The school adopted an automated and fully integrated self-service portal that enabled quick, easy implementation and immediate return on investment.

With out-of-the-box management and automation in support of its growing private cloud service, Aston University provided branded self-service capabilities and automated billing to customers, while increasing quality of service, improving administration effectiveness and reducing errors.

The journey to the private cloud varies somewhat for each organization that pursues that destination, but the general phases each faces along the way are consistent. By approaching these four stages with full knowledge of the challenges and best practices required to progress, companies are more likely to reach their private cloud goals and reap the rewards of a fully evolved virtualization model.