Storage and SQL Server capacity planning and configuration (SharePoint Server 2010)

 

Applies to: SharePoint Server 2010, SharePoint Foundation 2010

This article describes how to plan for and configure the storage and Microsoft SQL Server database tier in a Microsoft SharePoint Server 2010 environment.

The capacity planning information in this document provides guidelines for you to use in your planning. It is based on testing performed at Microsoft on live properties. However, your results may vary based on the equipment you use and the features and functionality that you implement for your sites.

Because SharePoint Server often runs in environments in which databases are managed by separate SQL Server database administrators, this document is intended for joint use by SharePoint Server farm implementers and SQL Server database administrators. It assumes significant understanding of both SharePoint Server and SQL Server.

This article assumes that you are familiar with the concepts presented in Capacity management and sizing for SharePoint Server 2010.

Design and configuration process for SharePoint 2010 Products storage and database tier

We recommend that you break the storage and database tier design process into the following steps. Each section provides detailed information about each design step, including storage requirements and best practices:

  • Gather storage and SQL Server space and I/O requirements

  • Choose SQL Server version and edition

  • Design storage architecture based on capacity and IO requirements

  • Estimate memory requirements

  • Understand network topology requirements

  • Configure SQL Server

  • Validate storage performance and reliability

Gather storage and SQL Server space and I/O requirements

Several SharePoint Server 2010 architectural factors influence storage design. The amount of content, features and service applications used, number of farms, and availability needs are key factors.

Before you start to plan storage, you should understand the databases that SharePoint Server 2010 can use.

In this section:

  • Databases used by SharePoint 2010 Products

  • Understand SQL Server and IOPS

  • Estimate core storage and IOPS needs

  • Determine service application storage and IOPs needs

  • Determine availability needs

Databases used by SharePoint 2010 Products

The databases that are installed with SharePoint Server 2010 depend on the features that are being used in the environment. All SharePoint 2010 Products environments rely on the SQL Server system databases. This section provides a summary of the databases installed with SharePoint Server 2010. For detailed information, see Database types and descriptions (SharePoint Server 2010) and Database model (https://go.microsoft.com/fwlink/p/?LinkId=187968).

Product version and edition Databases

SharePoint Foundation 2010

Configuration

Central Administration content

Content (one or more)

Usage and Health Data Collection

Business Data Connectivity

Application Registry service (if upgrading from Microsoft Office SharePoint Server 2007 Business Data Catalog)

Subscription Settings service (if it is enabled through Windows PowerShell)

Additional databases for SharePoint Server 2010 Standard edition

Search service application:

  • Search administration

  • Crawl (one or more)

  • Property (one or more)

User Profile service application:

  • Profile

  • Synchronization

  • Social tagging

Web analytics service application

  • Staging

  • Reporting

Secure store

State

Managed Metadata

Word Automation services

Additional databases for SharePoint Server 2010 Enterprise edition

PerformancePoint

Additional databases for Project Server 2010

Draft

Published

Archive

Reporting

Additional database for FAST Search Server

Search administration

If you are integrating more fully with SQL Server, your environment may also include additional databases, as in the following scenarios:

  • Microsoft SQL Server 2008 R2 PowerPivot for Microsoft SharePoint 2010 can be used in a SharePoint Server 2010 environment that includes SQL Server 2008 R2 Enterprise Edition and SQL Server Analysis Services. If in use, you must also plan to support the PowerPivot Application database, and the additional load on the system. For more information, see Plan a PowerPivot deployment in a SharePoint farm (https://go.microsoft.com/fwlink/p/?LinkID=186698).

  • The Microsoft SQL Server 2008 Reporting Services (SSRS) plug-in can be used with any SharePoint 2010 Products environment. If you are using the plug-in, plan to support the two SQL Server 2008 Reporting Services databases and the additional load that is required for SQL Server 2008 Reporting Services.

Understand SQL Server and IOPS

On any server that hosts SQL Server, it is very important that the server achieve the fastest response possible from the I/O subsystem.

More and faster disks or arrays provide sufficient I/O operations per second (IOPS) while maintaining low latency and queuing on all disks.

Slow response from the I/O subsystem cannot be compensated for by adding other types of resources such as CPU or memory; however, it can influence and cause issues throughout the farm. Plan for minimal latency before deployment, and monitor your existing systems.

Before you deploy a new farm, we recommend that you benchmark the I/O subsystem by using the SQLIO disk subsystem benchmark tool. For details, see SQLIO Disk Subsystem Benchmark Tool (https://go.microsoft.com/fwlink/p/?LinkID=105586).

For detailed information about how to analyze IOPS requirements from a SQL Server perspective, see Analyzing I/O Characteristics and Sizing Storage Systems for SQL Server Database Applications (http://sqlcat.com/whitepapers/archive/2010/05/10/analyzing-i-o-characteristics-and-sizing-storage-systems-for-sql-server-database-applications.aspx).

Estimate core storage and IOPS needs

Configuration and content storage and IOPs are the base layer that you must plan for in every SharePoint Server 2010 deployment.

Configuration storage and IOPS

Storage requirements for the Configuration database and the Central Administration content database are not large. We recommend that you allocate 2 GB for the Configuration database and 1 GB for the Central Administration content database. Over time, the Configuration database may grow beyond 1 GB, but it does not grow quickly — it grows by approximately 40 MB for each 50,000 site collections.

Transaction logs for the Configuration database can be large, therefore we recommend that you change the recovery model for the database from full to simple.

Note

If you want to use SQL Server database mirroring to provide availability for the Configuration database, you must use the full recovery model.

IOPS requirements for the Configuration database and Central Administration content database are minimal.

Content storage and IOPS

Estimating the storage and IOPS required for content databases is not a precise activity. In testing and explaining the following information, we intend to help you derive estimates to use for determining the initial size of your deployment. However, when your environment is running, we expect that you will revisit your capacity needs based on the data from your live environment.

For more information about our overall capacity planning methodology, see Capacity management and sizing for SharePoint Server 2010.

Estimate content database storage

The following process describes how to approximately estimate the storage required for content databases, without considering log files:

  1. Calculate the expected number of documents. This value is referred to as D in the formula.

    How you calculate the number of documents will be determined by the features that you are using. For example, for My Sites or collaboration sites, we recommend that you calculate the expected number of documents per user and multiply by the number of users. For records management or content publishing sites, you may calculate the number of documents that are managed and generated by a process.

    If you are migrating from a current system, it may be easier to extrapolate your current growth rate and usage. If you are creating a new system, review your existing file shares or other repositories and estimate based on that usage rate.

  2. Estimate the average size of the documents that you will be storing. This value is referred to as S in the formula. It may be worthwhile to estimate averages for different types or groups of sites. The average file size for My Sites, media repositories, and different department portals can vary significantly.

  3. Estimate the number of list items in the environment. This value is referred to as L in the formula.

    List items are more difficult to estimate than documents. We generally use an estimate of three times the number of documents (D), but this will vary based on how you expect to use your sites.

  4. Determine the approximate number of versions. Estimate the average number of versions any document in a library will have (this value will usually be much lower than the maximum allowed number of versions). This value is referred to as V in the formula.

    The value of V must be above zero.

  5. Use the following formula to estimate the size of your content databases:

    Database size = ((D × V) × S) + (10 KB × (L + (V × D)))

    The value of 10 KB in the formula is a constant that roughly estimates the amount of metadata required by SharePoint Server 2010. If your system requires significant use of metadata, you may want to increase this constant.

As an example, if you were to use the formula to estimate the amount of storage space required for the data files for a content database in a collaboration environment with the following characteristics, you would need approximately 105 GB.

Input Value

Number of documents (D)

200,000

Calculated by assuming 10,000 users times 20 documents

Average size of documents (S)

250 KB

List items (L)

600,000

Number of non-current versions (V)

2

Assuming that the maximum versions allowed is 10

Database size = (((200,000 x 2)) × 250) + ((10 KB × (600,000 + (200,000 x 2))) = 110,000,000 KB or 105 GB

Features that influence the size of content databases

The use of the following SharePoint Server 2010 features can significantly affect the size of content databases:

  • Recycle bins   Until a document is fully deleted from both the first stage and second stage recycle bin, it occupies space in a content database. Calculate how many documents are deleted each month to determine the effect of recycle bins on the size of content databases. For more information, see Configure Recycle Bin settings (SharePoint Server 2010).

  • Auditing   Audit data can quickly compound and use large amounts of space in a content database, especially if view auditing is turned on. Rather than letting audit data grow without restraint, we recommend that you only enable auditing on the events that are important to meet regulatory needs or internal controls. Use the following guidelines to estimate the space you will need to reserve for auditing data:

    • Estimate the number of new auditing entries for a site, and multiply this number by 2 KB (entries generally are limited to 4 KB, with an average size of about 1 KB).

    • Based on the space that you want to allocate, determine the number of days of audit logs you want to keep.

  • Office Web Apps. If Office Web Apps are being used, the Office Web Apps cache can significantly affect the size of a content database. By default, the Office Web Apps cache is configured to be 100 GB. For more information about the size of the Office Web Apps cache, see Manage the Office Web Apps cache.

Estimate content database IOPS requirements

IOPS requirements for content databases vary significantly based on how your environment is being used, and how much disk space and how many servers you have. In general, we recommend that you compare the predicted workload in your environment to one of the solutions that we tested. For more information, see Performance and capacity test results and recommendations (SharePoint Server 2010).

Important

The testing for the content in this section is not yet complete. Check back for additional information.

Estimate service application storage needs and IOPS

After estimating content storage and IOPs needs, you must next determine the storage and IOPs required by the service applications that are being used in your environment.

SharePoint Server 2010 service application storage and IOPs requirements

To estimate the storage requirements for the service applications in the system, you must first be aware of the service applications and how you will use them. Service applications that are available in SharePoint Server 2010 that have databases are listed in the following table.

Service application Size estimation recommendation

Search

Search requires three databases. Your environment may include multiple Property and Crawl databases.

The Search administration database is typically small: allocate 10 GB.

To estimate the required storage for your Property and Crawl databases, use the following multipliers:

  • Crawl: 0.046 × (sum of content databases)

  • Property: 0.015 × (sum of content databases)

The IOPS requirements for Search are significant.

  • For the Crawl database, search requires from 3,500 to 7,000 IOPS.

  • For the Property database, search requires 2,000 IOPS.

For detailed information about how to estimate capacity required for Search, see Performance and capacity test results and recommendations (SharePoint Server 2010).

FAST Search Server 2010 for SharePoint has a different architecture. The Crawl database has the same IOPS requirements, but the Property database is only used for people search and there is an additional Search administration database. For detailed information about FAST Search Server 2010 for SharePoint, see Plan search topology (FAST Search Server 2010 for SharePoint) and Performance and capacity management (FAST Search Server 2010 for SharePoint).

User Profile

The User Profile service application is associated with three databases: Profile, Synch, and Social Tagging.

To estimate the required storage for the databases, use the following information:

  • Profile. With default settings, in an environment configured to use Active Directory, the profile database requires approximately 1 MB per user profile.

  • Synchronization. With default settings, in an environment that has few groups per user, the synch database requires approximately 630 KB per user profile. 90% of the space will be used by the data file.

  • Social tagging. With default settings, the social tagging database requires approximately 0.009 MB per tag, comment, or rating. To estimate how many tags and notes users will create, consider the following information about the site del.icio.us:

    • Approximately 10% of users are considered active.

    • Active users create 4.5 tags and 1.8 comments per month.

In a live collaboration environment with 160,000 user profiles, 5 groups, 79,000 tags, comments and ratings (2,500 comments, 76,000 tags, and 800 ratings), and default settings, we saw the following sizes for these databases:

 

Database name Database size

Profile

155 GB

Synchronization

96 GB

Social tagging

0.66 GB

Managed metadata

The Managed Metadata service application has one database. The size of the database is affected by the number of content types and keywords used in the system. Many environments will include multiple instances of the Managed Metadata service application.

Web Analytics

Web Analytics has two databases: Staging and Reporting. Many factors influence the size of the databases. They include retention period, the daily volume of data being tracked, and the number of site collections, sites, and subsites in the Web application being analyzed. For detailed information about how to estimate their sizing and IOPS requirements, see Performance and capacity test results and recommendations (SharePoint Server 2010).

Secure store

The size of the Secure Store service application database is determined by the number of credentials in the store and the number of entries in the audit table. We recommend that you allocate 5 MB for each 1,000 credentials for it. It has minimal IOPS.

State

The State service application has one database. We recommend that you allocate 1 GB for it. It has minimal IOPS.

Word Automation service

The Word Automation service application has one database. We recommend that you allocate 1 GB for it. It has minimal IOPS.

PerformancePoint

The PerformancePoint service application has one database. We recommend that you allocate 1 GB for it. It has minimal IOPS.

Determine availability needs

Availability is the degree to which a SharePoint Server 2010 environment is perceived by users to be available. An available system is a system that is resilient — that is, incidents that affect service occur infrequently, and timely and effective action is taken when they do occur.

Availability requirements can significantly increase your storage needs. For detailed information, see Plan for availability (SharePoint Server 2010).

Choose SQL Server version and edition

Although SharePoint 2010 Products can run on Microsoft SQL Server 2008 R2, SQL Server 2008, or SQL Server 2005, we strongly recommend that you consider running your environment on the Enterprise Edition of SQL Server 2008 or SQL Server 2008 R2 to take advantage of the additional performance, availability, security, and management capabilities that it provides. For more information about the benefits of using SQL Server 2008 R2 Enterprise Edition, see SQL Server 2008 R2 and SharePoint 2010 Products: Better Together (white paper) (SharePoint Server 2010).

In particular, you should consider your need for the following features:

  • Backup compression   Backup compression can speed up any SharePoint backup, and is available in SQL Server 2008 Enterprise Edition or SQL Server 2008 R2 Standard edition. By setting the compression option in your backup script, or by configuring the server that is running SQL Server to compress by default, you can significantly reduce the size of your database backups and shipped logs. For more information, see Backup Compression (SQL Server) (https://go.microsoft.com/fwlink/p/?LinkId=129381&clcid=0x409).

    Note

    SQL Server data compression is not supported for SharePoint 2010 Products, except for the Search service application databases.

  • Transparent data encryption   If your security requirements include the need for transparent data encryption, you must use SQL Server Enterprise Edition.

  • Web Analytics service application   If you plan to use the Web Analytics service application for significant analysis, consider SQL Server Enterprise Edition so that the system can take advantage of table partitioning.

  • Content deployment   If you plan to use the content deployment feature, consider SQL Server Enterprise Edition so that the system can take advantage of SQL Server database snapshots.

  • Remote BLOB storage   If you want to take advantage of remote BLOB storage to a database or location outside the files associated with each content database, you must use SQL Server 2008 or SQL Server 2008 R2 Enterprise Edition.

  • Resource governor   Resource Governor is a technology introduced in SQL Server 2008 that enables you to manage SQL Server workloads and resources by specifying limits on resource consumption by incoming requests. Resource Governor enables you to differentiate workloads and allocate CPU and memory as they are requested, based on the limits that you specify. It is available only in SQL Server 2008 or SQL Server 2008 R2 Enterprise edition. For more information about using Resource Governor, see Managing SQL Server Workloads with Resource Governor.

    We recommend that you use Resource Governor with SharePoint Server 2010 to:

    • Limit the amount of SQL Server resources that the Web servers targeted by the search crawl component consume. As a best practice, we recommend limiting the crawl component to 10 percent CPU when the system is under load.

    • Monitor how many resources are consumed by each database in the system — for example, you can use Resource Governor to help you determine the best placement of databases among computers that are running SQL Server.

  • PowerPivot for SharePoint 2010   Enables users to share and collaborate on user-generated data models and analysis in Excel and in the browser while automatically refreshing those analyses. It is part of SQL Server 2008 R2 Enterprise Edition Analysis Services.

Design storage architecture based on capacity and I/O requirements

The storage architecture and disk types that you select for your environment can affect system performance.

In this section:

  • Choose a storage architecture

  • Choose disk types

  • Choose RAID types

Choose a storage architecture

Direct Attached Storage (DAS), Storage Area Network (SAN), and Network Attached Storage (NAS) storage architectures are supported with SharePoint Server 2010, although NAS is only supported for use with content databases that are configured to use remote BLOB storage. Your choice depends on factors within your business solution and your existing infrastructure.

Any storage architecture must support your availability needs and perform adequately in IOPS and latency. To be supported, the system must consistently return the first byte of data within 20 milliseconds (ms).

Direct Attached Storage (DAS)

DAS is a digital storage system that is directly attached to a server or workstation, without a storage network in between. DAS physical disk types include Serial Attached SCSI (SAS) and Serial Attached ATA (SATA).

In general, we recommend that you choose a DAS architecture when a shared storage platform cannot guarantee a response time of 20 ms and sufficient capacity for average and peak IOPs.

Storage Area Network (SAN)

SAN is an architecture to attach remote computer storage devices (such as disk arrays and tape libraries) to servers in such a way that the devices appear as locally attached to the operating system (for example, block storage).

In general, we recommend that you choose a SAN when the benefits of shared storage are important to your organization.

The benefits of shared storage include the following:

  • Easier to reallocate disk storage between servers.

  • Can serve multiple servers.

  • No limitations on the number of disks that can be accessed.

Network Attached Storage (NAS)

A NAS unit is a self-contained computer that is connected to a network. Its sole purpose is to supply file-based data storage services to other devices on the network. The operating system and other software on the NAS unit provide the functionality of data storage, file systems, and access to files, and the management of these functionalities (for example, file storage).

Note

NAS is only supported for use with content databases that are configured to use remote BLOB storage. Any network storage architecture must respond to a ping within 1 ms and must return the first byte of data within 20 ms. This restriction does not apply to the local SQL Server FILESTREAM provider, because it only stores data locally on the same server.

Choose disk types

The disk types that you use in the system can affect reliability and performance. All else being equal, larger drives increase mean seek time. SharePoint Server 2010 supports the following types of drives:

  • Small Computer System Interface (SCSI)

  • Serial Advanced Technology Attachment (SATA)

  • Serial-attached SCSI (SAS)

  • Fibre Channel (FC)

  • Integrated Device Electronics (IDE)

  • Solid State Drive (SSD) or Flash Disk

Choose RAID types

RAID (Redundant Array of Independent Disks) is often used to both improve the performance characteristics of individual disks (by striping data across several disks) and to provide protection from individual disk failures.

All RAID types are supported for SharePoint Server 2010; however, we recommend that you use RAID 10 or a vendor-specific RAID solution that has equivalent performance.

When you configure a RAID array, make sure that you align the file system to the offset that is supplied by the vendor. In the absence of vendor guidance, refer to SQL Server Predeployment I/O Best Practices (https://go.microsoft.com/fwlink/p/?LinkID=105583).

For more information about provisioning RAID and the SQL Server I/O subsystem, see SQL Server Best Practices Article (https://go.microsoft.com/fwlink/p/?LinkId=168612).

Estimate memory requirements

The memory required for SharePoint Server 2010 is directly related to the size of the content databases that you are hosting on a server that is running SQL Server.

As you add service applications and features, your requirements are likely to increase. The following table gives guidelines for the amount of memory we recommend.

Note

Our definitions of small and medium deployments are those described in the Reference architectures section of the article Capacity management and sizing overview for SharePoint Server 2010.

Combined size of content databases RAM recommended for computer running SQL Server

Minimum for small production deployments

8 GB

Minimum for medium production deployments

16 GB

Recommendation for up to 2 terabytes

32 GB

Recommendation for the range of 2 terabytes to 5 terabytes

64 GB

Recommendation for more than 5 terabytes

Additional RAM over 64 GB can improve SQL Server caching speed

Note

These values are higher than those recommended as the minimum values for SQL Server because of the distribution of data required for a SharePoint Server 2010 environment. For more information about SQL Server system requirements, see Hardware and Software Requirements for Installing SQL Server 2008 (https://go.microsoft.com/fwlink/p/?LinkId=129377).

Other factors that may influence the memory required include the following:

  • The use of SQL Server mirroring.

  • The frequent use of files larger than 15 megabytes (MB).

Understand network topology requirements

Plan the network connections within and between farms. We recommend that you use a network that has low latency.

The following list provides some best practices and recommendations:

  • All servers in the farm should have LAN bandwidth and latency to the server that is running SQL Server. Latency should be no greater than 1ms.

  • We do not recommend a wide area network (WAN) topology in which a server that is running SQL Server is deployed remotely from other components of the farm over a network that has latency greater than 1 ms. This topology has not been tested.

  • Plan for an adequate WAN network if you are planning to use SQL Server mirroring or log shipping to keep a remote site up-to-date.

  • We recommend that Web servers and application servers have two network adapters: one network adapter to handle end user traffic and the other to handle communication with the servers running SQL Server.

Configure SQL Server

The following sections describe how to plan to configure SQL Server for SharePoint Server 2010.

In this section:

  • Determine how many instances or servers are required

  • Configure storage and memory

  • Set SQL Server options

  • Configure databases

Estimate how many servers are required

In general, SharePoint Server 2010 was designed to take advantage of SQL Server scale out — that is, SharePoint Server 2010 may perform better with a large number of medium-size servers that are running SQL Server than with only a few large servers.

Always put SQL Server on a dedicated server that is not running any other farm roles or hosting databases for any other application, unless you are deploying the system on a stand-alone server.

The following is general guidance for when to deploy an additional server that will run SQL Server:

  • Add an additional database server when you have more than four Web servers that are running at full capacity.

  • Add an additional database server when your current server has reached its effective resource limits of RAM, CPU, disk IO throughput, disk capacity, or network throughput.

Note

Microsoft supports server configurations that do not follow this guidance.

To promote secure credential storage when you are running the Secure Store service application, we recommend that the secure store database be hosted on a separate database instance where access is limited to one administrator.

Configure storage and memory

On the server that is running SQL Server 2008, we recommend that the L2 cache per CPU have a minimum of 2 MB to improve memory.

Follow vendor storage configuration recommendations

For optimal performance when you configure a physical storage array, adhere to the hardware configuration recommendations supplied by the storage vendor instead of relying on the default values of the operating system.

If you do not have guidance from your vendor, we recommend that you use the DiskPart.exe disk configuration utility to configure storage for SQL Server 2008. For more information, see Predeployment I/O Best Practices (https://go.microsoft.com/fwlink/p/?LinkID=105583&clcid=0x409).

Provide as many resources as possible

Ensure that the SQL Server I/O channels to the disks are not shared by other applications, such as the paging file and Internet Information Services (IIS) logs.

Provide as much bus bandwidth as possible. Greater bus bandwidth helps improve reliability and performance. Consider that the disk is not the only user of bus bandwidth — for example, you must also account for network access.

Set SQL Server options

The following SQL Server settings and options should be configured before you deploy SharePoint Server.

  • Do not enable auto-create statistics on a SQL Server that is supporting SharePoint Server. SharePoint Server configures the required settings upon provisioning and upgrade. . Auto-create statistics can significantly change the execution plan of a query from one instance of SQL Server to another instance of SQL Server. Therefore, to provide consistent support for all customers, SharePoint Server provides coded hints for queries as needed to provide the best performance across all scenarios.

  • To ensure optimal performance, we strongly recommend that you set max degree of parallelism (MAXDOP) to 1 SQL Server instances that host SharePoint Server 2010 databases. For more information about how to set max degree of parallelism, see max degree of parallelism Option (https://go.microsoft.com/fwlink/p/?LinkId=189030).

  • To improve ease of maintenance, configure SQL Server connection aliases for each database server in your farm. A connection alias is an alternative name that can be used to connect to an instance of SQL Server. For more information, see How to: Set a SQL Server Alias (SQL Server Management Studio) (https://go.microsoft.com/fwlink/p/?LinkId=132064&clcid=0x409).

Configure databases

The following guidance describes best practices to plan for as you configure each database in your environment.

Separate and prioritize your data among disks

Ideally, you should place the tempdb database, content databases, Usage database, search databases, and SQL Server 2008 transaction logs on separate physical hard disks.

The following list provides some best practices and recommendations for prioritizing data:

  • When you prioritize data among faster disks, use the following ranking:

    1. Tempdb data files and transaction logs

    2. Database transaction log files

    3. Search databases, except for the Search administration database

    4. Database data files

    In a heavily read-oriented portal site, prioritize data over logs.

  • Testing and customer data show that SharePoint Server 2010 farm performance can be significantly impeded by insufficient disk I/O for tempdb. To avoid this issue, allocate dedicated disks for tempdb. If a high workload is projected or monitored — that is, the average read operation or the average write operation requires more than 20 ms — you might have to ease the bottleneck by either separating the files across disks or by replacing the disks with faster disks.

  • For best performance, place the tempdb on a RAID 10 array. The number of tempdb data files should equal the number of core CPUs, and the tempdb data files should be set at an equal size. Count dual core processors as two CPUs for this purpose. Count each processor that supports hyper-threading as a single CPU. For more information, see Optimizing tempdb Performance (https://go.microsoft.com/fwlink/p/?LinkID=148537).

  • Separate database data and transaction log files across different disks. If files must share disks because the files are too small to warrant a whole disk or stripe, or you have a shortage of disk space, put files that have different usage patterns on the same disk to minimize simultaneous access requests.

  • Consult your storage hardware vendor for information about how to configure all logs and the search databases for write optimization for your particular storage solution.

Use multiple data files for content databases

Follow these recommendations for best performance:

  • Only create files in the primary filegroup for the database.

  • Distribute the files across separate disks.

  • The number of data files should be less than or equal to the number of core CPUs. Count dual core processors as two CPUs for this purpose. Count each processor that supports hyper-threading as a single CPU.

  • Create data files of equal size.

Important

Although you can use the backup and recovery tools that are built in to SharePoint Server 2010 to back up and recover multiple data files, if you overwrite in the same location, the tools cannot restore multiple data files to a different location. For this reason, we strongly recommend that when you use multiple data files for a content database, you use SQL Server backup and recovery tools. For more information about how to back up and recover SharePoint Server 2010, see Plan for backup and recovery in SharePoint Server 2010.

For more information about how to create and manage filegroups, see Physical Database Files and Filegroups (https://go.microsoft.com/fwlink/p/?LinkId=117909).

Limit content database size to improve manageability

Plan for database sizing that will improve manageability, performance, and ease of upgrade for your environment.

To help ensure system performance, we recommended that you limit the size of content databases to 200 GB, except when specific usage scenarios and conditions support larger sizes. For more information about content database size limits, see the Content database limits section in SharePoint Server 2010 capacity management: Software boundaries and limits.

We generally recommend that a site collection should not exceed 100 GB unless it is the only site collection in the database so that you can use the SharePoint Server 2010 granular backup tools to move a site collection to another database if you need to.

For more information about large-scale document repositories, see "Estimate Performance and Capacity Requirements for Large Scale Document Repositories, available from Performance and capacity test results and recommendations (SharePoint Server 2010).

Proactively manage the growth of data and log files

We recommend that you proactively manage the growth of data and log files by considering the following recommendations:

  • As much as possible, pre-grow all data and log files to their anticipated final size.

  • We recommend that you enable autogrowth for safety reasons. Do not rely on the default autogrowth settings. Consider the following guidelines when configuring autogrowth:

    • When you plan content databases that exceed the recommended size (200 GB), set the database autogrowth value to a fixed number of megabytes instead of to a percentage. This will reduce the frequency with which SQL Server increases the size of a file. Increasing file size is a blocking operation that involves filling the new space with empty pages.

    • Set the autogrowth value for the Search service application Property Store database to 10 percent.

    • If the calculated size of the content database is not expected to reach the recommended maximum size of 200 GB within the next year, set it to the maximum size the database is predicted to reach within a year — with 20 percent additional margin for error — by using the ALTER DATABASE MAXSIZE property. Periodically review this setting to make sure it is still an appropriate value based on past growth rates.

  • Maintain a level of at least 25 percent available space across disks to allow for growth and peak usage patterns. If you are managing growth by adding disks to a RAID array or allocating more storage, monitor disk size closely to avoid running out of space.

Validate and monitor storage and SQL Server performance

Test that your performance and backup solution on your hardware enables you to meet your service level agreements (SLAs). In particular, test the I/O subsystem of the computer that is running SQL Server to ensure that performance is satisfactory.

Test the backup solution that you are using to ensure that it can back up the system within the available maintenance window. If the backup solution cannot meet the SLAs your business requires, consider using an incremental backup solution such as System Center Data Protection Manager (DPM) 2010.

It is important to track the following resource components of a server that is running SQL Server: CPU, memory, cache/hit ratio, and I/O subsystem. When one or more of the components seems slow or overburdened, analyze the appropriate strategy based on the current and projected workload. For more information, see Troubleshooting Performance Problems in SQL Server 2008 (https://go.microsoft.com/fwlink/p/?LinkID=168448).

The following section lists the performance counters that we recommend that you use to monitor the performance of the SQL Server databases that are running in your SharePoint Server 2010 environment. Also listed are approximate healthy values for each counter.

For details about how to monitor performance and use performance counters, see Monitoring Performance (https://go.microsoft.com/fwlink/p/?LinkId=189032).

SQL Server counters to monitor

Monitor the following SQL Server counters to ensure the health of your servers:

  • General statistics   This object provides counters to monitor general server-wide activity, such as the number of current connections and the number of users connecting and disconnecting per second from computers running an instance of SQL Server. Consider monitoring the following counter:

    • User connections   This counter shows the amount of user connections on your computer running SQL Server. If you see this number rise by 500 percent from your baseline, you may see a performance reduction.
  • Databases   This object provides counters to monitor bulk copy operations, backup and restore throughput, and transaction log activities. Monitor transactions and the transaction log to determine how much user activity is occurring in the database and how full the transaction log is becoming. The amount of user activity can determine the performance of the database and affect log size, locking, and replication. Monitoring low-level log activity to gauge user activity and resource usage can help you to identify performance bottlenecks. Consider monitoring the following counter:

    • Transactions/sec   This counter shows the amount of transactions on a given database or on the entire server per second. This number is more for your baseline and to help you troubleshoot issues.
  • Locks   This object provides information about SQL Server locks on individual resource types. Consider monitoring the following counters:

    • Average Wait Time (ms)   This counter shows the average amount of wait time for each lock request that resulted in a wait.

    • Lock Wait Time (ms)   This counter shows the wait time for locks in the last second.

    • Lock waits/sec   This counter shows the number of locks per second that could not be satisfied immediately and had to wait for resources.

    • Number of deadlocks/sec   This counter shows the number of deadlocks on the computer running SQL Server per second. This should not rise above 0.

  • Latches   This object provides counters to monitor internal SQL Server resource locks called latches. Monitoring the latches to determine user activity and resource usage can help you to identify performance bottlenecks. Consider monitoring the following counters:

    • Average Latch Wait Time (ms)   This counter shows the average latch wait time for latch requests that had to wait.

    • Latch Waits/sec   This counter shows the number of latch requests that could not be granted immediately.

  • SQL Statistics   This object provides counters to monitor compilation and the type of requests sent to an instance of SQL Server. Monitoring the number of query compilations and recompilations and the number of batches received by an instance of SQL Server gives you an indication of how quickly SQL Server is processing user queries and how effectively the query optimizer is processing the queries. Consider monitoring the following counters:

    • SQL Compilations/sec   This counter indicates the number of times the compile code path is entered per second.

    • SQL Re-Compilations/sec   This counter indicates the number statement recompiles per second.

  • Buffer Manager   This object provides counters to monitor how SQL Server uses memory to store data pages, internal data structures, and the procedure cache, as well as counters to monitor the physical I/O as SQL Server reads and writes database pages. Consider monitoring the following counter:

    • Buffer Cache Hit Ratio

    • This counter shows the percentage of pages that were found in the buffer cache without having to read from disk. The ratio is the total number of cache hits divided by the total number of cache lookups over the last few thousand page accesses. Because reading from the cache is much less expensive than reading from disk, you want this ratio to be high. Generally, you can increase the buffer cache hit ratio by increasing the amount of memory available to SQL Server.

  • Plan Cache   This object provides counters to monitor how SQL Server uses memory to store objects such as stored procedures, ad hoc and prepared Transact-SQL statements, and triggers. Consider monitoring the following counter:

    • Cache Hit Ratio

    • This counter indicates the ratio between cache hits and lookups for plans.

Physical server counters to monitor

Monitor the following counters to ensure the health of your computers running SQL Server:

  • Processor: % Processor Time: _Total   This counter shows the percentage of time that the processor is executing application or operating system processes other than Idle. On the computer that is running SQL Server, this counter should be kept between 50 percent and 75 percent. In case of constant overloading, investigate whether there is abnormal process activity or if the server needs additional CPUs.

  • System: Processor Queue Length   This counter shows the number of threads in the processor queue. Monitor this counter to ensure that it remains less than two times the number of core CPUs.

  • Memory: Available Mbytes   This counter shows the amount of physical memory, in megabytes, available to processes running on the computer. Monitor this counter to ensure that you maintain a level of at least 20 percent of the total available physical RAM.

  • Memory: Pages/sec   This counter shows the rate at which pages are read from or written to disk to resolve hard page faults. Monitor this counter to ensure that it remains under 100.

For more information and memory troubleshooting methods, see SQL Server 2005 Monitoring Memory Usage (https://go.microsoft.com/fwlink/p/?LinkID=105585).

Disk counters to monitor

Monitor the following counters to ensure the health of disks. Note that the following values represent values measured over time — not values that occur during a sudden spike and not values that are based on a single measurement.

  • Physical Disk: % Disk Time: DataDrive   This counter shows the percentage of elapsed time that the selected disk drive is busy servicing read or write requests–it is a general indicator of how busy the disk is. If the PhysicalDisk: % Disk Time counter is high (more than 90 percent), check the PhysicalDisk: Current Disk Queue Length counter to see how many system requests are waiting for disk access. The number of waiting I/O requests should be sustained at no more than 1.5 to 2 times the number of spindles that make up the physical disk.

  • Logical Disk: Disk Transfers/sec   This counter shows the rate at which read and write operations are performed on the disk. Use this counter to monitor growth trends and forecast appropriately.

  • Logical Disk: Disk Read Bytes/sec and Logical Disk: Disk Write Bytes/sec   These counters show the rate at which bytes are transferred from the disk during read or write operations.

  • Logical Disk: Avg. Disk Bytes/Read   This counter shows the average number of bytes transferred from the disk during read operations. This value can reflect disk latency — larger read operations can result in slightly increased latency.

  • Logical Disk: Avg. Disk Bytes/Write   This counter shows the average number of bytes transferred to the disk during write operations. This value can reflect disk latency — larger write operations can result in slightly increased latency.

  • Logical Disk: Current Disk Queue Length   This counter shows the number of requests outstanding on the disk at the time that the performance data is collected. For this counter, lower values are better. Values greater than 2 per disk may indicate a bottleneck and should be investigated. This means that a value of up to 8 may be acceptable for a logical unit (LUN) made up of 4 disks. Bottlenecks can create a backlog that can spread beyond the current server that is accessing the disk and result in long wait times for users. Possible solutions to a bottleneck are to add more disks to the RAID array, replace existing disks with faster disks, or move some data to other disks.

  • Logical Disk: Avg. Disk Queue Length   This counter shows the average number of both read and write requests that were queued for the selected disk during the sample interval. The rule is that there should be two or fewer outstanding read and write requests per spindle, but this can be difficult to measure because of storage virtualization and differences in RAID levels between configurations. Look for larger than average disk queue lengths in combination with larger than average disk latencies. This combination can indicate that the storage array cache is being overused or that spindle sharing with other applications is affecting performance.

  • Logical Disk: Avg. Disk sec/Read and Logical Disk: Avg. Disk sec/Write   These counters show the average time, in seconds, of a read or write operation to the disk. Monitor these counters to ensure that they remain below 85 percent of the disk capacity. Disk access time increases exponentially if read or write operations are more than 85 percent of disk capacity. To determine the specific capacity for your hardware, refer to the vendor documentation or use the SQLIO Disk Subsystem Benchmark Tool to calculate it. For more information, see SQLIO Disk Subsystem Benchmark Tool (https://go.microsoft.com/fwlink/p/?LinkID=105586).

    • Logical Disk: Avg. Disk sec/Read   This counter shows the average time, in seconds, of a read operation from the disk. On a well-tuned system, ideal values are from 1 through 5 ms for logs (ideally 1 ms on a cached array), and from 4 through 20 ms for data (ideally less than 10 ms). Higher latencies can occur during peak times, but if high values occur regularly, you should investigate the cause.

    • Logical Disk: Avg. Disk sec/Write   This counter shows the average time, in seconds, of a write operation to the disk. On a well-tuned system, ideal values are from 1 through 5 ms for logs (ideally 1 ms on a cached array), and from 4 through 20 ms for data (ideally less than 10 ms). Higher latencies can occur during peak times, but if high values occur regularly, you should investigate the cause.

    When you are using RAID configurations with the **Logical Disk: Avg. Disk Bytes/Read   ** or Logical Disk: Avg. Disk Bytes/Write counters, use the formulas listed in the following table to determine the rate of input and output on the disk.

    RAID level Formula

    RAID 0

    I/Os per disk = (reads + writes) / number of disks

    RAID 1

    I/Os per disk = [reads + (2 × writes)] / 2

    RAID 5

    I/Os per disk = [reads + (4 × writes)] / number of disks

    RAID 10

    I/Os per disk = [reads + (2 × writes)] / number of disks

    For example, if you have a RAID 1 system that has two physical disks, and your counters are at the values that are shown in the following table.

    Counter Value

    Avg. Disk sec/Read

    80

    Logical Disk: Avg. Disk sec/Write

    70

    Avg. Disk Queue Length

    5

    The I/O value per disk can be calculated as follows: (80 + (2 × 70))/2 = 110

    The disk queue length can be calculated as follows: 5/2 = 2.5

    In this situation, you have a borderline I/O bottleneck.

Other monitoring tools

You can also monitor disk latency and analyze trends by using the sys.dm_io_virtual_file_stats dynamic management view in SQL Server 2008. For more information, see sys.dm_io_virtual_file_stats (Transact-SQL) (https://go.microsoft.com/fwlink/p/?LinkID=105587).

See Also

Other Resources

Resource Center: Capacity Management for SharePoint Server 2010
Resource Center: SQL Server and SharePoint Server 2010 Databases