Large Scale SharePoint SQL Deployments

Considerations for large-scale SharePoint deployments on SQL Server
Name: Joel Oleson
Title: Sr. Tech Prod Mgr
Company: Quest Software

Audience Poll
New to SharePoint?
SQL Admins?
Large-scale Implementation (>1TB) experience?
Scalability or performance issues in SharePoint deployments?

Session Overview
Lightweight
Understanding SharePoint databases
SQL Performance
SQL Server 2008 with SharePoint
Heavyweight
Architectural Design Considerations
Real-world scenarios
Business Requirements
Logical and Physical Architecture
Architectural Design Statistical Results
Appendix: DB Sizes, Content Distribution…

=
Lightweight

Real World Examples
Information based on real-world, large-scale SharePoint Implementations.
Large software company (Microsoft)
Intranet Portal for 120K users
Global Enterprise Collaboration Solution (~20TB)
Scalable Hosting Solution (SharePoint Online)
Large automotive manufacturer
Loan Origination Application / Document Repository
~50 Million content items (~6 TB)

Understanding the SharePoint Databases

Disk I/O Demand
Most Demand
Medium Demand
Low Demand
*Content..
Search
Config
Temp
Model
+SSP
Master
Tlogs
* Except during backup and Indexing + Except during Profile Import

Top Performance Killers
Indexing/Crawling
Backup (SQL & Tape)
Profile Import
Misc Timer Jobs – User Sync for large #s of Users
STSADM Backup/Restore
Large List Operations
Heavy User Operation List Import/Write

Content Db

Config

SQL Server 2008 with Windows Server 2008
Transactional Performance with SQL Server 2008 Dramatically outperformed SQL 2005 on Win 2003.
Compressed backup in the box
Support for SQL External Blob Storage
Increased resiliency
Transparent Encryption
See Performance Gains athttp://msdn.microsoft.com/en-us/library/dd263442.aspx

=
heavyweight

Database Volumes
Separate database volumes into unique LUN’s consisting of unique physical disk spindles.
In a heavily read-oriented internet (portal) site, prioritize data over logs.
Separate out Search database transaction log from content database transaction logs.

SQL TempDB Data Files
Optimal TempDB data file sizes can be calculated using the following formula:
[MAX DB SIZE (KB)] X [.25] / [# CORES] = DATA FILE SIZE (KB)
Calculation result (starting size) should be roughly equal to 25% of the largest content or search DB.
Use RAID 10; separate LUN from other database objects (content, search, etc…).
“Autogrow” feature set to a fixed amount; if auto grow occurs, permanently increase TempDB size.
TempDB Log file separated to unique LUN.

Content Databases
100 content databases per Web application
100GB per content database
CAUTION: DB locking issues reported in collaborative DM scenarios above 100GB
Need to ensure that you understand the issues based on number of users, usage profiles, etc…
Service Level Agreement (SLA) requirements for backup and restore will also have an impact on this decision.

Content Databases - Continued
Pre-construct and pre-size
Use RAID 5 or RAID 10 logical units
RAID 10 is the best choice when cost is not a concern.
RAID 5 will be sufficient and will save on costs, since content databases tend to be more read intensive than write intensive.
Multi-core computer running SQL Server
Primary file group could consist of a data file for each CPU core present in SQL Server.

Database Maintenance
SQL Server SP2 is needed if using the DB maintenance wizard (KB930887).
Plan regular defrag of databases
Performance - Average Disk Queue Length
Single Digit values are optimal.
Occasional double-digit values aren’t a large concern.
Sustained triple-digit values require attention.

Performance
The recommended practice for separating the database volume types for the transaction log files to unique LUN’s follows.
Content Database Log Files.
Search Database Log Files.
Consider filegroups for search database

Topology
A single list should not have more than 2,000 items per list container.
A container represents the root of the list, as well as any folders within the list; a folder is a container because other list items are stored within it.
Whitepaper: Working with large lists in Office SharePoint Server 2007 (Steve Peschka)
http://go.microsoft.com/fwlink/?LinkId=95450
Disk Drive Speed
15K RPM recommended.
IIS Application Pools
Ensure “Max Used Memory” setting utilizes all the available RAM in your WFE’s.

STSADM Command-line Tool and CreateSiteInNewDBOperation
Gary Lapoint STSADM Extensions for Site Collection DB maintenance
Codeplex.com/governance tools for archive & delete capture

Large Scale Manufacture

Real-world Scenarios
Automotive Mfgr. Business Requirements (Phase I)
Loan Origination Application built on Office SharePoint Server 2007
Ability to manage10.5 million images.
System performance with a “normal” input load defined as receipt of 27,000 images per business day = 10 hours.
Simulate user load to represent 200 users for search, view & update with 2x peak

Data Load Process (Phase I)
Used KnowledgeLake Document Release Engine
Loaded 9.17 documents/second per server
Employs a high-volume, storage-based folder architecture within SharePoint to ensure UI responsiveness.
Executed on 4 servers. Using this application, we were able to achieve:
An average document load throughput of 36.6 documents per second!
An average daily input of 3.17 million documents!
10.5 million documents with only 28% utilization!

Data Load Process (Phase II)
15 million documents consisting of Word (.docx), Excel (.xlsx), PowerPoint (.pptx) and Adobe PDF.
Five Web Front-Ends were used for the load process.
Peak Load Rate:
24.3 docs per second/2.1 million documents per day.
Average Load Rate:
~1.9 million documents per day.
Load Time:
8 days.NOTE: Load rates included automation process that created the PDF files.

What does the logical architecture look like?!

What does the physical architecture look like?!
Scale OUT…
Scale UP…

What does the site topology look like?!
Phase I
17 Divisional Site Collections / DB’s
Phase II
10 Departmental Site Collections / DB’s

What does the storage architecture look like?

Database SizesPhase I

Architectural Design Statistical ResultsPhase I
Designed Once / Built Once
No architecture OR configuration changes were required after the initial build was completed.
10.5+ million documents loaded into the system in approximately 60 hours!
Full Crawl indexed 10 Million items in 32 hours!
Average content database size for divisional breakouts was 60GB

Architectural Design Statistical ResultsPhase II
Search database size was 539GB.
Lesson Learned: Large search database caused disk I/O contention; break this out into multiple data file allocations matching the number of core processors on SQL Server, and spread them over unique LUN’s.
Total Index size was 162GB!
Average Content database size for Divisional breakouts was 200.65GB!
Average Content database size for Departmental breakouts was 137.60GB!

Large Scale Pharma

Pharmaceutical Business Requirements
Collaboration Portal built on Office SharePoint Server 2007
Validate ~40TB of content storage.
Identify performance characteristics and provide guidance around content database sizing
FAST search integration

Data Load Process
71,524,357 documents loaded across two SharePoint Farms 10.92 days!
Content was spread across the farms into 165 unique content databases.
6,240 Site Collections, each containing 10 sub-sites for a total of 62,400 sites.
Database sizes were pre-configured to vary in size from 100GB to 350GB to determine performance and/or SLA impacts.

What does the logical architecture look like?!

What does the physical architecture look like?!

What does the site topology look like?!
165 Content DB’s
6,240 Site Collections
10 Sub-Sites in each collection:
62,400 Sites!

What does the storage architecture look like?

Architectural Design Statistical ResultsConclusion
User Loads
Stress tests included 2 - 3,000 concurrent users.
Based on the 10% rule, testing completed equated to an environment representing 300,000 users!
RAWnumber of RPS during peak times is 1,469 at Pharma.
773 RPS, which equates to 346.59 ACTUAL RPS!
FAST Search Integration
Successfully integrated FAST search capabilities, indexed content corpus and served search results as expected.

Large-Scale Case Study Available
SharePoint Scalability and Performance Whitepaper
Contains majority of content you will see here, along with test results you won’t see here.
TechNet topic: http://go.microsoft.com/fwlink/?LinkId=120901
Word 2007 format: http://go.microsoft.com/fwlink/?LinkId=120881
Word 2000-2003 format: http://go.microsoft.com/fwlink/?LinkId=120890
PDF format: http://go.microsoft.com/fwlink/?LinkId=120891

question & answer

Appendix

Database SizesMPSC/Nissan Phase I

Database SizesMPSC/Nissan Phase II

Performance of Components Over Time
MPSC/Nissan Phase I
14 individual performance tests were run to simulate various load scenarios.

How do we pull all this together?!
PharmaContent Database Distribution
Substitute “F1” with SQL Server number to generate unique DB’s
Farm 1: 2 SQL
Farm 2: 1 SQL
165 Content Databases!

How do we pull all this together?!
PharmaData Load Statistics

Architectural Design Statistical ResultsTesting Results – 300GB Content Databases

Required Slide
© 2009 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.
The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

http://blog.van-huizen.com	18
http://www.slideshare.net	12

Large Scale SharePoint SQL Deployments

by Joel Oleson on Jul 04, 2009

Accessibility

Categories

Tags

Upload Details

Usage Rights

2 Embeds 30

Statistics

Large Scale SharePoint SQL Deployments — Presentation Transcript