People and process challenges can be a bigger hindrance to virtual machine implementation than technical issues (see the ITA Premium research note, “Five Pitfalls of Virtualization and How to Avoid Them”). Business and operational assessments focus on people and process issues of virtualization.
- A business assessment focuses externally on addressing the general business benefits of virtualization (how it saves money) as well as the specific needs and perspectives of individual business units.
- An operation assessment focuses internally on the operational capability of the IT department to manage a virtual infrastructure.
The two assessments are being considered together for this research note because they are closely related. For example, if the IT department has formalized repeatable procedures for server configuration and performance management, it can more evidently meet the service level requirements of a given business unit.
Steps in the business and operational readiness assessment include:
- Establish “team virtualization.”
- Review top level goals and devise success measures.
- Assess the capital and operating expense of physical server replacement.
- Assess the virtualization readiness of the application users/owners.
- Assess the capability maturity for managing virtual infrastructure.
- Avoid poison pills: relate assessment to virtualization candidates’ list.
- Develop a checklist of key success indicators, and measure pilot implementation.
The purpose of an assessment is to analyze the current state of affairs and decide what needs to be done to move to a desired state. The key deliverables of the business and operational assessments will be:
- Documentation of business unit requirements and position on virtualization.
- Appreciation of risk tolerance across the enterprise.
- A list of business and operational success indicators for the virtualization implementation.
A virtual server implementation is best seen within the context of planning and deploying a new kind of foundational architecture for x86 server management. A previous ITA Premium research note in this series, “Virtual Servers Planning Goals: Take a Baby Bear Approach,” argues that the best approach is one that combines larger planning for a virtual machine architecture with careful and incremental adoption.
With the implementation of a new technology, such as server virtualization, the focus can often fall mainly on technical issues. The key question is: “What will it take to make it work?” Technical assessment for virtualization focuses on figuring out whether and how a physical server will work as a virtual machine. However, technical glitches are not the only issues that can frustrate a virtualization implementation.
People and process issues can stop a virtualization implementation project as surely as a technical failure. Some of the non-technical hurdles include:
- Application owners or key users refusing to have their applications virtualized because they don’t believe they will get comparable or better service, reliability, or operational flexibility.
- Failure of the IT department to secure the investment necessary to build appropriate virtual server architecture.
- Inability of the IT department to clearly demonstrate how security, availability, and reliability will be guaranteed for the virtual infrastructure.
Where the technical assessment helps to answer the question, “Will it work?” the business and operational assessments help to answer questions such as:
- Will it save the enterprise money?
- Will it improve business efficiency, security, and flexibility?
- Will it help the company execute on larger strategic goals?
- Will it help individual units achieve their business goals (or at least not get in the way by adding new complexities and risk)?
By helping to answer these questions, the business assessment and the operational assessment will be as crucial to successful implementation as the technology assessment. The three assessments should be seen as critical components of an overall assessment approach.
It’s About Audience and Trust
Virtualization of server infrastructure will have an impact beyond the scope of server administrators. It will impact network management, security management, storage management, and business continuity/disaster recovery management. Application owners also need to understand how their applications will function/perform in a virtual environment. Executive sponsors need to know how virtualization will save money and advance the enterprise toward strategic goals.
The goal of business and operational assessment is to be able to speak to these stakeholders in a language that they understand and to also build trust. A business stakeholder, for example, probably could care less about discussions concerning percentage of processor utilization, usage spikes, and storage Input/Output per second. For stakeholders, a simple yes or no answer to the question “Will it work?” will suffice.
What they are more likely to be interested in is an estimate of the capital and operational expense savings of virtual server consolidation. They will also want a discussion of what a business unit will gain in terms of service levels in return for giving up sole ownership of a physical server asset.
The trust factor relates directly to the IT department’s ability to relate mature, repeatable, and accountable management processes to a virtual environment. For example, in dealing with line-of-business stakeholders, providing service level guarantees for virtual machines will be easier if there is a system of performance monitoring for service level agreements already practiced. If meeting enterprise service level and application performance demands has been previously met only through the acquisition and configuration of new physical server hardware, the IT department has work to do.
Executive Sponsorship Is Critical
Executive sponsorship (i.e. having a senior level champion) is the single most important component of a successful virtualization strategy. Because the implications, and potential benefits, of virtualization go beyond any one project or group within the enterprise, top level sponsorship is critical. An executive sponsor will facilitate successful long term virtualization implementation in the following ways:
- Ensure funding for core virtualization infrastructure. A long term strategy of “virtualize unless otherwise” may require up-front investment that would be difficult to justify for a one-off virtualization of a few servers. For example, a storage area network (SAN) is critical for ensuring high availability and optimum performance of a virtual infrastructure. SANs are expensive. The cost of the SAN should be justified in context of the longer term goal that has high level support.
- Establish the high level goals and review progress. Part of the business and operational assessments will be to establish how virtualization will help the enterprise achieve broader goals. As the champion of these larger goals, the executive sponsor will help measure the success of virtualization against these goals.
- Drive change, and overcome internal barriers. While it would be preferable for all stakeholders in the enterprise to be enthusiastic about virtualization, internal resistance to change will likely be encountered. If making a value case for virtualization is the carrot, the authority of the executive sponsor is the stick.
The third role above is particularly important for managing resistance to change and overcoming risk aversion. Though a strong technical case for the capability, stability, and recoverability of virtual machines can be made, it should be recognized that this is a significant change in understanding how IT services will work.
“Because the boss says so” is not a good argument for change, in fact it isn’t an argument at all, but it is a valuable weapon to have in the arsenal if all other arguments fail.
Implementation & Integration
- Establish “Team Virtualization.” A previous ITA Premium research note in this series on virtualization implementation (“Virtual Servers Planning Goals: Take a Baby Bear Approach”) calls for a scope analysis to identify dependencies, opportunities, and stakeholders. If this scoping has been done, the virtualization implementer has a list of key stakeholders that have already been identified. These stakeholders will likely include:
- Representatives from business units that have applications which are potential virtualization candidates.
- IT staff that are responsible for disaster recovery, storage management, server management, and network management.
- Potential executive sponsor or sponsors for a virtualization initiative.
From this group of people, a core virtualization team should be established. Whether this team is a formal body or an informal communications circle will depend on the culture, size, and management maturity of the organization. The key here is to establish who should be in the loop for virtualization to succeed and devise a process for keeping them in the loop.
Team virtualization will participate in the business and operational assessments and, if pilot implementations meet expectations, become internal champions for server virtualization.
- Review top level goals and devise success measures. The “Virtual Servers Planning Goals: Take a Baby Bear Approach” research note also calls for identifying some high level goals to which virtualization can be tied. Take a look at these goals now in the more specific terms of how success might be measured. For example:
- If the enterprise has identified being green as a strategic goal, assess the current consumption of electricity for power and cooling of servers. Establish measures for gauging the potential per server reduction in this consumption if servers were virtual machines running on fewer physical server instances.
- If the enterprise has recently established recovery time and recovery point objectives for disaster recovery planning, assess the RTO and RPO of current physical server infrastructure. Establish how RTO and RPO will be tested and estimated for potential improvement on these goals when current servers are virtual servers in a high availability infrastructure.
These are just two examples. Individual results for this assessment will depend on the current business and IT objectives of the enterprise. This is where discussion with potential executive sponsor(s) will be critical.
Assess the capital and operating expense of physical server replacement. Consolidation ratio (i.e. the number of virtual machines that will be able to share a single physical host) is important to calculating savings in capital and operating expenses (capex and opex) related to virtual server deployment. Info-Tech, for example, has estimated that the break-even point on expenses of physical versus virtual servers occurs at a consolidation ratio of roughly 3 to 1.
A consolidation rate higher than 3:1 should mean capex and opex savings for the enterprise. Vendors often promise consolidation rates of 20:1 or more so even if actual implementation is lower there should still be justifiable savings. However, the 3:1 break-even is only an estimate based on typical per server costs and licensing fees. The enterprise needs to assess their own server needs and estimate actual per physical-server costs of replacement so that it can be compared to actual per virtual server costs.
If actual per physical-server costs are lower and per virtual-server costs are higher, then the break-even consolidation ratio will be higher. Factors that could drive per virtual-server costs higher include vendor licensing schemes that are not virtualization friendly as well as capital expenses for network storage.
The success metric for this assessment will be to exceed the break-even consolidation ratio that is realistic for the enterprise.
- Assess the virtualization readiness of the application users/owners. Applications enable business processes. A business assessment for virtualization should include an assessment on how virtualization will impact applications. Discuss the requirements of business process owners and assess how their needs are currently being met by critical applications. Some issues to consider include:
- Current service levels guaranteed for critical applications including up-time and availability.
- Current turn-around time for requests for new server-based applications.
- Current cost and time requirements for development and testing for business-critical application upgrades.
- General level of risk aversion with business stakeholder.
From these assessments will come some baseline business requirements against which virtual servers and management of virtual infrastructure can be measured.
Assess the capability maturity for managing virtual infrastructure. As noted above, managing a virtual infrastructure may require more mature and accountable processes than what is currently practiced. If the enterprise is going to have an overall strategic goal of moving to ubiquitous virtualization, IT processes will need to improve.
Think in terms of the classic Capability Maturity Model (CMM). Originally developed as a way of measuring and refining an organization’s software development capabilities, the CMM can be broadly applied to the management of any IT process. The five levels of the CMM are:
- Initial: Success depends on ad hoc efforts.
- Repeatable: Basic project management techniques are employed and processes are defined and documented.
- Defined: Organization has developed standardized, integrated, and documented management processes.
- Managed: Organization standardizes the monitoring and analysis of defined processes to maintain quality.
- Optimized: Building on current processes, innovations are introduced to improve service or quality.
Virtualization’s ultimate success depends on managing IT services as a utility to provision business needs. Utility management requires at least a maturity level of 3 – Defined. An IT department that focuses only on acquiring new physical servers for each business demand is likely operating at an Initial, ad hoc, level of 1.
Assess where the current capability maturity of IT for managing IT services is in a defined, standardized, and documented fashion. If level 3 management practice is present, what will be needed to adapt these processes to a virtual infrastructure?
Avoid poison pills: relate assessment to virtualization candidates list. Lack of executive sponsorship, application owner resistance, and lack of operational readiness should be red flags for virtual server implementation. Relate the results of the business and operational assessments to the virtualization candidates identified in the technical assessment.
The goal here is to avoid a poison pill server. A poison pill is a server that appears to be an easy candidate for virtualization from a technical perspective but is actually a political landmine. Poison pill and red flag servers could include:
- Servers that are used by business stakeholders who are resistant to virtualization.
- Servers that the executive sponsor does not believe should be virtualized.
- Critical servers for which the IT department can not prove its operational competency to manage for processes, such as service level guarantee, disaster recovery, or security.
The ITA Premium “Assessing Candidate Servers for Virtualization” tool can be employed to compile the results of the technical, business, and operational assessment in support of virtualization decision making. It includes fields for relating a given server candidate to possible business and operation impediments to virtualization.
Develop key success indicators and measure pilot implementation. Based on the above assessment work, Team Virtualization should be able to develop a list of key success indicators for virtual server implementations. These can be compiled into a checklist or score card to monitor the success of the project and report to an executive sponsor.
It was noted above that the initial implementation of virtual servers should be seen as a use case and pilot to inform the planning process. Info-Tech has established, for example, that in a typical 100 server shop scenario, initial implementation is likely part of a hardware refresh project. Virtualization is being considered because a portion of the server infrastructure – maybe 20 – has reached end-of-life and needs to be replaced.
Though the short-term goal may be to refresh the hardware for 20 end-of-life servers, this project should not be treated as a one-off acquisition but as a pilot project and the first step toward building a managed virtual infrastructure. Using the measures agreed upon in this business and operational assessment, the 20 server virtualization project can be measured both for short term immediate return on investment and for what it teaches the enterprise for longer term continued virtual machine deployment.
Business and operational assessments are as critical for virtual server implementation as is the technical assessment that identifies virtualization candidates. People and process issues are more likely to be an impediment to implementation than technical limitations. Identify and resolve these issues through assessments that establish baselines of business and operational measures for virtualization success.
© 1998-2010 Info-Tech Research Group. All rights reserved. Reprinted by permission
Info-Tech's products and services combine actionable insight and relevant advice with ready-to-use tools and templates that cover the full spectrum of IT concerns. For more information, go to www.infotech.com.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access