Departmental collaboration environment technical case study (SharePoint Server 2010)

 

Gilt für: SharePoint Server 2010

Letztes Änderungsdatum des Themas: 2016-11-30

This document describes a specific deployment of Microsoft SharePoint Server 2010. It includes the following:

• Technical case study environment specifications, such as hardware, farm topology and configuration

• The workload that includes the number, and types, of users or clients, and environment usage characteristics

• Technical case study farm dataset that includes database contents and Search indexes

• Health and performance data that is specific to the environment

In this article:

• Prerequisite information

• Introduction to this environment

• Specifications

• Workload

• Dataset

• Health and Performance Data

Prerequisite information

Before reading this document, make sure that you understand the key concepts behind SharePoint Server 2010 capacity management. The following documentation will help you learn about the recommended approach to capacity management and provide context for helping you understand how to make effective use of the information in this document, and also define the terms used throughout this document.

For more conceptual information about performance and capacity that you might find valuable in understanding the context of the data in this technical case study, see the following documents:

• Kapazitätsverwaltung und Größengestaltung für SharePoint Server 2010 (Übersicht)

• SharePoint Server 2010-Kapazitätsverwaltung: Softwarebeschränkungen und -grenzen

Introduction to this environment

This white paper describes an actual SharePoint Server 2010 environment at Microsoft. Use this document to compare with your planned workload and usage characteristics. If your planned design is similar, you can use the deployment described here as a starting point for your own installation.

This document includes the following:

• Specifications, which include hardware, topology and configuration

• Workload, which is the demand on the farm that includes the number of users, and the usage characteristics

• Dataset that includes database sizes

• Health and performance data that is specific to the environment

This document is part of a series of Performance and capacity technical case studies (SharePoint Server 2010) about SharePoint environments at Microsoft.

The SharePoint Server 2010 environment described in this document is a production environment at a large, geographically distributed company. Employees use this environment to track projects, collaborate on documents, and share information within their department. This environment is also used for internal testing, and is frequently upgraded to the latest SharePoint Server pre-release versions as they become available.

As many as 9,000 unique users visit the environment on a busy day, generating up to 470 requests per second (RPS) during peak hours. Because this is an intranet site, all users are authenticated.

The information that is provided in this document reflects the departmental collaboration environment on a typical day.

Specifications

This section provides detailed information about the hardware, software, topology, and configuration of the case-study environment.

Hardware

This section provides details about the server computers that were used in this environment.

Hinweis

This environment is scaled to accommodate pre-release builds of SharePoint Server 2010 and other products. Hence, the hardware deployed has larger capacity than necessary to serve the demand typically experienced by this environment. This hardware is described only to provide additional context for this environment and serve as a starting point for similar environments.
It is important to conduct your own capacity management based on your planned workload and usage characteristics. For more information about the capacity management process, see Kapazitätsverwaltung und Größengestaltung für SharePoint Server 2010 (Übersicht).

Web Servers

There are four Web servers in the farm, each with identical hardware. Three serve content, and the fourth is a dedicated search crawl target.

Web Server	WFE1-2	WFE3-4
Processor(s)	2 quad core @ 2.33 GHz	2 quad core @ 2.33 GHz
RAM	32 GB	16 GB
Operating system	Windows Server 2008, 64 bit	Windows Server 2008, 64 bit
Size of the SharePoint drive	3x146GB 15K SAS (3 RAID 1 Disks) Disk 1: OS Disk 2: Swap and BLOB Cache Disk 3: Logs and Temp directory	3x146GB 15K SAS (3 RAID 1 Disks) Disk 1: OS Disk 2: Swap and BLOB Cache Disk 3: Logs and Temp directory
Number of network adapters	2	2
Network adapter speed	1 Gigabit	1 Gigabit
Authentication	Windows NTLM	Windows NTLM
Load balancer type	Hardware load balancing	Hardware load balancing
Software version	SharePoint Server 2010 (pre-release version)	SharePoint Server 2010 (pre-release version)
Services running locally	Search Query	WFE3 – No services WFE4 – Search crawl target

Application Server

There are four application servers in the farm.

Web Server	APP1-3	APP4
Processor(s)	2 quad core @ 2.33 GHz	2 quad core @ 2.33 GHz
RAM	16 GB	16 GB
Operating system	Windows Server 2008, 64 bit	Windows Server 2008, 64 bit
Size of the SharePoint drive	3x146GB 15K SAS (3 RAID 1 Disks) Disk 1: OS Disk 2: Swap and BLOB Cache Disk 3: Logs and Temp directory	2x136GB 15K SAS (RAID 0) 4x60GB SSD, SATA (RAID 5) Disk 1: OS Disk 2: Swap and BLOB Cache Disk 3: Logs and Temp directory
Number of network adapters	2	2
Network adapter speed	1 Gigabit	1 Gigabit
Authentication	Windows NTLM	Windows NTLM
Load balancer type	Hardware load balancing	Hardware load balancing
Software version	SharePoint Server 2010 (pre-release version)	SharePoint Server 2010 (pre-release version)
Services running locally	APP1 – Central Administration and all applications except for Office Web Applications APP2 – All applications (including Office Web Applications) APP3 – Office Web Applications	Search Crawler

Database Servers

There are three database servers, one running the default SQL Server instance housing the content databases, one running the Usage and Web Analytics databases, and one running the Search databases.

Database	DB1 – Default Instance	DB2	DB3
Processor(s)	4 quad core @ 3.2 GHz	2 quad core @ 3.2 GHz	2 quad core @ 3.2 GHz
RAM	32 GB	16 GB	32 GB
Operating system	Windows Server 2008 SP1, 64 bit	Windows Server 2008 SP1, 64 bit	Windows Server 2008 SP1, 64 bit
Storage and geometry	5x146GB 15K SAS + SAN Disk 1: OS (2 disk RAID 10) Disk 2: Swap (2 disk RAID 10) Disk 3: Direct Attached Storage (16 disk RAID 10, Temp DB data) SAS 146 GB 15K Disk 4: Direct Attached Storage (16 disk RAID 10, Temp DB data) SAS 146 GB 15K Disk 5-15: SAN using fiber connection. When possible, one database per two disks. Separating logs and data between LUNs. 15K drives.	6x450GB 15K SAS Directly attached 14x146GB 15K SAS Disk 1: Usage logs and OS Disk 2: Usage data	2x136GB 15K SAS (RAID 0) 6x60GB SSD, SATA (RAID 5) Disk 1: OS Disk 2: Swap and BLOB Cache Disk 3: Logs and Temp directory. Solid state drives. 6-60GB Solid state drives (RAID 5)
Number of network adapters	2	2	2
Network adapter speed	1 Gigabit	1 Gigabit	1 Gigabit
Authentication	Windows NTLM	Windows NTLM	Windows NTLM
Software version	SQL Server 2008	SQL Server 2008	SQL Server 2008 R2

Topology

The following diagram shows the topology for this farm.

Configuration

The following table enumerates settings that were made that affect performance or capacity in the environment.

Setting	Value	Notes
Site collection: Object Caching (On | Off) Anonymous Cache Profile (select) Anonymous Cache Profile (select) Object Cache (Off | n MB) Cross List Query Cache Changes (Every Time | Every n seconds)	On Disabled Disabled On – 100GB 60 seconds	Enabling the output cache improves server efficiency by reducing calls to the database for data that is frequently requested.
Site collection cache profile (select)	Intranet (Collaboration Site)	“Allow writers to view cached content” is checked, bypassing the ordinary behavior of not letting people with edit permissions to have their pages cached.
Object Cache (Off | n MB)	On – 500 MB	The default is 100 MB. Increasing this setting enables additional data to be stored in the front-end Web server memory.
Usage Service: Trace Log – days to store log files (default: 14 days)	5 days	The default is 14 days. Lowering this setting can save disk space on the server where the log files are stored.
Query Logging Threshold: Microsoft SharePoint Foundation Database – configure QueryLoggingThreshold to 1 second	1 second	The default is 5 seconds. Lowering this setting can save bandwidth and CPU on the database server.
Database Server – Default Instance: Max degree of parallelism	1	The default is 0. To ensure optimal performance, we strongly recommend that you set max degree of parallelism to 1 for database servers 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).

Workload

This section describes the workload, which is the demand on the farm that includes the number of users, and the usage characteristics.

Workload Characteristics	Value
Average Requests per Second (RPS)	165
Average RPS at peak time (11 AM-3 PM)	216
Total number of unique users per day	9186
Average concurrent users	189
Maximum concurrent users	322
Total # of requests per day	7,124,943

This table shows the number of requests for each user agent.

User Agent	Requests	Percentage of Total
Search (crawl)	4,373,433	67.61%
Outlook	897,183	13.87%
OneNote	456,917	7.06%
DAV	273,391	4.23%
Browser	247,303	3.82%
Word	94,465	1.46%
SharePoint Workspaces	70,651	1.09%
Office Web Applications	45,125	0.70%
Excel	8,826	0.14%
Access	1,698	0.03%

Dataset

This section describes the case study farm dataset that includes database sizes and Search indexes.

Dataset Characteristics	Value
Database size (combined)	1.8 TB
BLOB size	1.68 TB
Number of content databases	18
Total number of databases	36
Number of site collections	7,499
Number of Web applications	7
Number of sites	42,457
Search index size (number of items)	4.6 million

Health and Performance Data

This section provides health and performance data that is specific to the case study environment.

General Counters

Metric	Value
Availability (uptime)	99.9995%
Failure Rate	0.0005%
Average memory used	0.89 GB
Maximum memory used	5.13 GB
Search Crawl % of Traffic (Search client requests / total requests)	82.5%

The following charts show the average CPU utilization and latency for this environment:

In this document, latency is divided into four categories. The 50th percentile latency is typically used to measure the server’s responsiveness. It means that half of the requests are served within that response time. The 95th percentile latency is typically used to measure spikes in server response times. It means that 95% of requests are served within that response time, and therefore, 5% of the requests experience slower response times.

Database Counters

Metric	Value
Average Disk queue length	1.42
Disk Queue Length: Reads	1.38
Disk Queue Length: Writes	0.04
Disk Reads/sec	56.51
Disk Writes/sec	17.60
SQL Compilations/second	13.11
SQL Re-compilations/second	0.14
SQL Locks: Average Wait Time	294.56 ms
SQL Locks: Lock Wait Time	867.53 ms
SQL Locks: Deadlocks Per Second	1.87
SQL Latches: Average Wait Time	5.10 ms
SQL Cache Hit Ratio	99.77%

See Also

Other Resources

Resource Center: Capacity Management for SharePoint Server 2010