Plan for virtualization (SharePoint Foundation 2010)
Applies to: SharePoint Foundation 2010
This article describes the planning process to follow in order to successfully deploy Microsoft SharePoint Foundation 2010 in a virtual environment. Each step in the planning process includes links to the appropriate documentation. It is assumed that you have determined the SharePoint Foundation 2010 solution that you want to deploy in a virtual environment. On the surface, deploying a SharePoint Foundation 2010 farm on virtual machines is the same as deploying a farm on physical servers. However, deploying in a virtual environment involves a different level of planning that takes into account the characteristics of Windows Server 2008 Hyper-V technology as well as how virtual machines, the virtual network adapters, and virtual hard disks are implemented on a virtualization server.
Before you start developing your virtualization plan, we recommend that you read the Hyper-V Planning and Deployment Guide (https://go.microsoft.com/fwlink/p/?LinkId=187964).
Detailed information about the following subjects is out-of-scope for this article, but is provided in other articles:
Capacity management
Security requirements
Health and performance monitoring
Backup and recovery
A virtual environment consists of two interrelated layers, one physical and one virtual. A configuration change in either layer effects servers in the other layer. This interrelationship becomes evident when you plan for, deploy, and use SharePoint Foundation 2010 in a virtual environment.
Create a plan for deploying SharePoint Foundation 2010 in a virtual environment
You should approach planning for a virtual farm the same way as you would plan for a physical farm. Most, if not all, of the issues and requirements for deploying SharePoint Foundation 2010 on physical servers apply equally to virtual machines. Any decisions that you make, such as minimum processor or memory requirements, have a direct bearing on number of virtualization hosts required, as well as their ability to adequately support the virtual machines that you identify for the farm.
After you finish planning a physical farm, you have all the information you need to design virtualization 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 lifecycle. In fact, you may determine that some farm server roles are not good candidates for virtualization.
The key planning steps, tasks, and references are summarized in following the procedure.
To create a virtualization plan
Determine virtualization scope
Determining the scope of farm virtualization is a key contributing factor to successfully implementing, managing, and evaluating your virtualization project. When determining scope, you have to decide whether you will virtualize some or the entire supporting virtual machine infrastructure.
Use the following list of tasks to determine the scope of virtualization.
Task 1: Identify all the farms that are required to implement your solution. Take into consideration the fact that most solutions have several farm components. For example, an Internet-facing Web portal typically has a publishing farm, an authoring farm, and a testing or quality assurance farm.
Task 2: For each farm, determine the number of servers that are required as well as the role that each server will have in the farm.
Task 3: Identify which farms you want to deploy in in a virtual environment.
Refining the scope of a solution also refines the scope of a deployment, which makes it easier to implement and manage. For more information, see Site and solution planning (SharePoint Foundation 2010). In many cases, solutions share common elements; however, each solution may have its own requirements. For more information, see Fundamental site planning (SharePoint Foundation 2010). The article shows one of the popular solutions.
Note
Expect to refine the scope of your solution as you move through the phases of deploying your farm in a production environment.
Identify servers to virtualize
Identify servers that are good candidates for virtualization. From a technical and Microsoft support perspective, all SharePoint servers can be virtualized. The decision to virtualize a particular farm server should be based on:
Corporate compliance policies (for example, legal and technical)
Benefits derived from server consolidation, such as reduced power consumption and physical space requirements. For more information, see Server virtualization (https://go.microsoft.com/fwlink/p/?LinkId=187965).
Capacity requirements (see next planning step)
Identify capacity requirements for each farm server
Determine the resource requirements for each farm server as if it was a physical server. Take into account specialized server roles, such as hosting Enterprise Search components. You need to specify the amount of resources needed for each of the following server components:
Memory
Number of processors and minimum clock speed
Number and size of hard disks
Number of network adapters and their required throughput speed
Determine if virtual machine can meet physical requirements.
You have to determine whether each virtual machine that you identified in Step 3 can meet the capacity requirements of a corresponding physical server. At a minimum, complete the following tasks:
Task 1: Assess the memory requirement in the context of available virtualization host capacity.
Task 2: Assess the processor requirement. Hyper-V has a hard limit of four virtual processors per virtual machine. If a physical farm server requires eight processors, determine whether this requirement can be met by scaling out the number of virtual machines in a farm.
Task 3: Assess the virtual machine storage requirement in the context of local physical storage or SAN.
Determine virtualization host requirements
Determine the minimum host requirements (memory, number of cores, number and size of local hard drives, number of network adapters)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
Depending on the manufacturer and computer model, you may not be able to increase capacity. You need to have this information before you use or purchase a server.
Extra host capacity: Determine whether or not the host has the capacity to scale up existing virtual machines, or to add additional virtual machines. This is very important if you plan to use Hyper-V failover clustering, quick migration, or live migration.
Important
Plan for peak load and determine how short term spikes in load will be handled.
Design virtualization architecture
A well-designed architecture is required for a successful solution. For SharePoint Foundation 2010, a basic three-tier topology provides the foundation for all the solutions. The following elements form a good design that is based on the recommended foundation topology:
Good overall performance
Ease of maintenance and upgrade
Flexibility
Scalability
High availability
A virtualization architecture model consists of the virtualization hosts and the virtual machines that make up the farm topology. This model enables you to visualize the virtual environment that you plan to deploy.
Note
Be prepared to refine the architecture as you move through the planning process. The following steps may dictate changes to the architecture.
Identify storage requirements
Determine how much local physical storage or SAN storage is required for Hyper-V-related storage such as configuration files, Virtual Hard Disks (VHDs), and snapshots.
Identify backup and recovery requirements
In addition to the farm servers, you have to plan backup and recovery for all or part of a farm. For more information, see Backup and recovery (SharePoint Foundation 2010).
Determine high availability requirements and design a solution
Identify approaches for achieving high availability for Web servers, application servers, and databases. Typical strategies include the following:
Redundant hardware and servers
Hot-swappable components
Failover clustering for virtual and physical servers. For more information, see Hyper-V: Using Hyper-V and Failover Clustering (https://go.microsoft.com/fwlink/p/?LinkId=187967).
Clustering or mirroring for database servers. For more information, see Plan for availability (SharePoint Foundation 2010).
Identify health and capacity indicators for monitoring the virtual environment.
Combine the key indicators that you derived in the previous steps with the planning you did for SharePoint Foundation 2010. For more information, see Server farm and environment planning (SharePoint Foundation 2010) . You have to determine all the health and capacity indicators in order to collect measurements from the following objects in the virtual environment:
Virtual machines with SharePoint Foundation 2010 installed
Virtual machines that are not part of the farm, such as a firewall server
Virtualization hosts
Network components
After you start to collect data from the virtual environment, you can create a baseline, which can be used to assess and tune the virtual environment during deployment and after the farm goes into production.
Create a deployment plan for the deployment phase of the system lifecycle.
For more information, see the SharePoint 2010 Products Deployment model, available in the Technical diagrams (SharePoint Foundation 2010) article.
Create a maintenance plan
Create a maintenance plan that enables you to implement password changes and apply software updates, service packs, and hotfixes. This plan should include the virtual machines and the virtualization hosts.
See Also
Other Resources
Resource Center: Virtualization for SharePoint Foundation 2010