Export (0) Print
Expand All

Create a virtualization plan for SharePoint 2013

Published: July 16, 2012

Summary: Learn about what to plan before you install, configure, and test SharePoint 2013 in a Windows Server 2008 environment.

Applies to:  SharePoint Foundation 2013 | SharePoint Server 2013 

A virtualization plan includes critical elements, key planning tasks, and decision points. Your plan's scope and level of detail determines the success of a SharePoint 2013 deployment in a test, pre-production, or production virtual environment.

In this article:

Introduction

In the real world, phases do not always depend on preceding phases. You can do many planning tasks concurrently. The major phases in this planning document are organized sequentially to show the logical flow of information gathering, decision-making, and plan creation. These phases and suggested sequence are not absolute; you can, and should, change the sequence and tasks to suit the needs of your organization. The goal is that the final plan contains all the information that you need to deploy SharePoint 2013 in a virtual environment.

Article scope

Although this article focuses on Windows Server 2008 Hyper-V technology, you can apply our planning concepts, guidance and tasks to any virtualization solution that the Server Virtualization Validation Program (SVVP) validates. For more information, see the Server Virtualization Program FAQ

The following subjects are mentioned in this article, but detailed information about each subject area is provided in separate articles.

  • Business continuity

  • Security

  • Performance and capacity

  • Maintenance and operations

Obtain reference material and tools

Reference material and tools that are relevant to the technologies that you plan to deploy are valuable as you plan. They reduce the possibility of missed steps, and templates or tools reduce errors and add to consistency. We recommend that you obtain copies of the following:

We recommend that you review the entire list of Solution Accelerators at Solution Accelerator Downloads to see if there are additional Solution Accelerators that you can use for your virtual farm.

Evaluate in-house resources and infrastructure

A critical first step when you create your plan is to inventory, document, and assess the availability of in-house resources and the technical infrastructure. The availability of adequate resources and infrastructure are determining factors when you plan how to budget, schedule, implement, maintain, and use SharePoint 2013 in a virtual environment.

If you do not have detailed specifications for the farm servers and virtualization server hosts, you can still derive a preliminary definition for system requirements. As you move forward in the planning process, you will refine the farm requirements and assess the available in-house resources and infrastructure again.

The minimum base requirements for a virtual farm are as follows:

  • Windows Server 2008 R2

  • Hyper-V

  • SharePoint 2013

  • SQL Server 2008 R2 or SQL Server 2012

  • A suitable hardware, storage, and network infrastructure to support the farm.

You have to verify the base technologies that are available to determine what you can reuse or what you have to buy for the virtualization project. In either case, you have to identify all the infrastructure deployment requirements and a schedule to have everything in place in time to deploy the farm. For the existing technologies, you have to evaluate the breadth and depth of in-house knowledge and experience that you can use to support the virtualization project. The presence or absence of these elements affects several aspects of your virtualization project, which include but is not limited to the following:

  • Implementation costs exclusive of hardware, software and licensing.

  • Human resource requirements.

  • Training requirements (schedule and cost) for users and the IT group.

  • Hardware sizing and costs, which you derive from existing or new benchmark data.

  • Pre-production phases and benchmark requirements.

  • The time will take to put the farm into full production.

After you finish your assessment you can then decide whether you can use in-house resources to handle all aspects of your virtual farm implementation. If external resources such as vendors or partners are required, identify the project tasks that you have to outsource.

You should also consider outsourcing other tasks to free up in-house resources who can focus on implementation. In the process, you can add to the in-house skillset and the organization's knowledge base. The use of outside resources changes the total project, and there may be other significant changes, such as changes to the schedule. If you decide that outsourcing is the best alternative for certain tasks, you have to incorporate the schedules of the external teams in the plan.

When you inventory and evaluate the infrastructure, you must also include existing IT policies, procedures, standards and best practices. A well-scoped infrastructure assessment highlights additional technical requirements such as hardware, software, and licenses, and also determines capabilities and limitations in the context of virtualization. A well-known example is virtualizing database servers, which is not permitted in many organizations, especially those that have a team that is dedicated to database support.

Create a project team

A project team is made up of suitable personnel who have the experience or skills that are required for different phases of the project. For example, technical members of the team collect and analyze benchmark data. However, non-technical team members who understand workflows between business units are very important for different aspects of user acceptance testing. In addition, team members aren't necessarily engaged in all the project phases. For example, a member of the legal team may only be engaged in establishing governance requirements for the farm. Representatives from different business units would only be involved in user acceptance testing for the SharePoint features that affect their teams.

A project team typically has the following members:

  • IT staff from the necessary product support areas (for example, database and networking) define initial benchmark criteria for the project. During the deployment phases they collect and analyze data, and validate/refine the benchmarks.

    note Note:

    User acceptance testing. As is the case with farm testing in a physical environment, users from each business unit that is affected by or will use the farm participate in user acceptance testing. The fact that the platform is virtualized should be transparent to the users.

  • Assuming that there is some degree of customization, members of the development team validate functionality and behavior and debug farm elements as the project moves through each deployment phase.

    note Note:

    It might be necessary to revalidate custom code that is now running on virtualized farm servers versus a solution that was designed to run, or was already running on physical servers.

You have to clearly lay out expectations, identify responsibilities, and obtain commitments before the project starts to guarantee that you have sufficient team and specific time commitments. Finally you must work with all the team members to establish acceptance criteria which will be used to determine when the farm can be put into production.

Identify virtual farm requirements

We recommend that you approach design of a virtual farm in the same way that you design a physical farm. Most issues and requirements to deploy on physical servers apply equally to virtual machines. Any decisions that you make, such as minimum processor or memory requirements, have a direct effect on the number of servers in the farm topology and corresponding virtualization host requirements. The host computers must have sufficient resources to support the virtual machines that you identify for the farm.

The following list shows some typical farm requirements that apply to physical and virtual farms.

  • Identify the required roles (for example, queries) for a specific SharePoint solution.

  • Decide how to distribute the SharePoint roles on farm servers.

  • Determine how many servers are needed for the farm.

  • Decide how each server needs to be configured.

The starting point to determine farm server specifications are the minimum hardware and software requirements for SharePoint 2013. For more information about these requirements (such as specific updates that you must install), see Hardware and software requirements for SharePoint 2013.

The next steps are to refine farm server specifications by using benchmark data from an existing SharePoint farm or from benchmark tests that are based on published performance and limits guidance.

Important Important:

If you use benchmark data from an existing farm, you still have to run benchmark tests.

Important Important:

Given the same configuration (number of processors, amount of memory, and hard disk size) the results of previous benchmark tests consistently show that a virtual machine has lower overall performance (for example, average response time, requests per second, and lower IOPS) than a physical computer. You must take these results into account when you design a virtual farm topology.

Determine virtualization scope and architecture

As you begin to design your architecture you have to decide whether to virtualize all the farm servers or have a mix of virtual machines and physical computers.

If your strategy is to implement a heterogeneous SharePoint virtualization environment, then you have to identify servers that are the best candidates for virtualization. From a technical and Microsoft support perspective, all servers that run SharePoint 2013 can be virtualized. The decision to virtualize a particular farm server should be based on specific requirements and considerations, such as the following:

  • Corporate compliance policies (for example, legal and technical) might not support virtualization, or virtualization host server and storage location.

  • Estimated performance and capacity requirements.

  • Security requirements.

  • Support and maintenance requirements

For more information about virtualization scope and architectures, see Overview of farm virtualization and architectures for SharePoint 2013.

After you create a topology design for a physical farm, you have a starting point that you can use to design a virtual architecture. Ideally, this architecture is as close as possible to the final virtualization solution that you intend to put into production. Realistically, the architecture is likely to change as you move through the deployment phase of the system life-cycle. For more information, see Detailed design and system specification process for a virtual SharePoint 2013 farm.

note Note:

As you develop benchmarks during the pilot and pre-production phases of deployment, you may decide that some server roles are not the best candidates for virtualization. If the scope of your virtualization is a homogeneous environment then you have to scale up or scale out to improve performance.

Identify virtual infrastructure requirements

Determine the minimum virtualization server host requirements (amount of memory, number of sockets (physical processors) and cores, number and size of local hard disks, number of network adapters).

Important Important:

The number of physical processors determines licensing cost for Windows Server virtualization. For more information, see Virtualization support and licensing in SharePoint 2013.

You should also consider and plan for the following:

  • Scalability

    Determine if you can add more CPUs, more memory, more hard disks, and more network adapters to the host computer.

    Important Important:

    The manufacturer, computer model, and hardware architecture might limit growth capability or even worse, may prevent an increase in capacity. You must know the limits of a potential host server before you try to reuse a server or purchase a new one.

  • Reserve capacity

    Determine whether the host has sufficient reserve capacity to scale up the existing virtual machines, or to scale out by adding virtual machines. This is very important if you plan to use Hyper-V failover clustering, quick migration, or live migration.

Important Important:

You must make sure that your strategy for responding to seasonal or random load increases on the virtual machines includes the host servers.

Identify virtualization management and maintenance requirements

Virtualization adds to the complexity of the IT environment and may add new challenges that your IT group must overcome to effectively manage the farm and its infrastructure. Frontline support is different, as is operations support and ongoing maintenance.

The virtual farm presents two additional levels of support requirements: the virtual machines and the virtualization host servers. Although the requirements of each level are similar to a farm that is deployed on physical servers, these requirements are not identical. There are enough unique elements to warrant an evaluation of the existing support infrastructure.

IT groups in most organizations have obtained or developed software tools and utilities for monitoring and managing their physical computing environment.

  • Third party tools and utilities - Most third-party software is customizable to a degree and IT groups customize tools to meet the changing automation requirements.

  • In house-tools and utilities - This software may only be a collection of batch files and scripts that have evolved in response to the changing environment but they do provide a degree of automation that does the job.

The challenge for the IT group is to determine whether any current tools can meet the demands of a virtual environment, and at what cost. Third-party tools may have reached the limits of customization or may not have caught up to virtualization requirements. Similarly, the in-house tool collection may not be more customizable, or the cost and time requirements may not be justifiable. Your analysis will determine whether you have to invest in new tools to manage the virtual farm.

If you decide to make an investment in tools to manage the new environment, we recommend Microsoft System Center 2012 it provides a robust set of tools that are well integrated with Hyper-V as well as SharePoint products. The System Center toolset includes the following programs that have broader capabilities in addition to managing your virtual environment.

  • Virtual Machine Manager (VMM) is a management solution for the virtualized datacenter. It enables you to configure and manage your virtualization host, networking, and storage resources to create and deploy virtual machines and services to private clouds.

  • System Center 2012 - Operations Manager provides infrastructure monitoring that is flexible and cost-effective. It helps ensure the predictable performance and availability of important applications, and it offers comprehensive monitoring for your datacenter and cloud, both private and public.

    note Note:

    Operations Manager also provides a System Center Management Pack for SharePoint Server 2013 and a System Center Management Pack for SharePoint Foundation 2013.

  • Microsoft System Center 2012 Configuration Manager helps you deliver user experiences across a wide range of devices, while also making sure that you meet corporate control and compliance requirements.

  • System Center 2012 - Data Protection Manager (DPM) enables disk-based and tape-based data protection and recovery for servers such as SQL Server, Exchange Server, SharePoint 2013, virtual servers, and file servers. DPM also supports Windows desktops and portable computers. DPM can also centrally manage system state and Bare Metal Recovery (BMR).

  • Orchestrator is a workflow management solution for the data center. Orchestrator lets you automate the creation, monitoring, and deployment of resources in your environment.

note Note:

Consider obtaining the Virtual Machine Servicing Tool from the Microsoft Download Center. This Solution Accelerator provides automated tools and guidance that IT professionals can use to update offline virtual machines, templates, and virtual hard disks efficiently and without exposing them to security risks.

Create a deployment plan

At this point in the planning process, you must create an implementation plan that describes the phases that move the virtual farm to production. The scope and details of your implementation plan is the same as if you were deploying a SharePoint products farm on physical servers.

Use the following table, which describes the phases for putting a physical farm into production as a guide to implement your virtual farm.

Implementation phases

Name Steps Output Characteristics

Proof of Concept (PoC)

Used to determine whether the solution will meet business needs and to determine the appropriate infrastructure.

  • Deploy farm

  • Deploy solution

  • Collect benchmark data

  • Evaluate proof of concept

  • Refine goals and infrastructure requirements

  • Updated solution plan

  • Updated topology

  • Updated resource requirements

PoCs are often created for evaluating a product or solution. A PoC can be hosted in a development environment or on a small production computer. Sometime s the PoC moves through several phases in continuous development.

Pilot (small scale test)

Use to test a solution on a small scale

Pilots are used to test solution readiness (no real data, just functional testing.) A pilot can also be used to test for production characteristics by using real data and performing actual work, which we recommend.

  • Deploy pilot farm

  • Deploy pilot solution

  • Collect benchmark data

  • Evaluate pilot

  • Refine goals and infrastructure requirements

  • Determine operations requirements

  • Updated solution plan

  • Updated topology and resource requirements

  • An operations plan

A limited set of users has access to the environment to test the functionality and performance of the solution and infrastructure.

User Acceptance Test (UAT)

The purpose is to provide a pre-production environment that is used for testing solutions against a subset or complete copy of production data.

This test is also used for validating backups or the operations procedures.

  • Deploy UAT farm

  • Deploy UAT solution

  • Implement operations plan

  • Evaluate solution

  • Evaluate operations plan

  • Test capacity and performance

Updated operations plan

The topology should be as similar to the production environment as possible.

Testers make sure that that all solution elements function as expected in a network and security conditions match those of the production environment.

Production

This is the live environment that that users interact with to do their jobs.

  • Deploy production farm

  • Deploy production solution

  • Implement operations plan

  • Deploy additional environments such as authoring and staging farms, and services farms

Deployed production farm

It is business-critical and all the appropriate service level agreements (SLAs) are in place.

For each of the previous phases you have to determine the answers to the following questions:

  • Who must be involved in the tests during each phase?

  • What must be tested?

  • What benchmark data is getting collected?

  • What are the criteria for moving to the next phase?

  • Who is responsible for signing off on the successful completion of the phase?

The duration of each phase should be determined by meeting predefined acceptance criteria in addition to any date-driven release goals. However, a time goal does provide a useful measure of the actual duration of a phase and validates time-based estimates. You can use the results for future deployments or upgrades.

Review and adjust the plan

Review and revise your plan before and during the implementation of the virtual farm in a production capacity. Keep the plan up to date -- it provides valuable information for a project post-mortem and can be used to plan future projects.

Was this page helpful?
(1500 characters remaining)
Thank you for your feedback
Show:
© 2014 Microsoft