IBM Fast Track DataWarehouse SQLServer 2012 Std x240 7TB ConfigurationGuide

Configuration guide for IBM Flex System x240 and IBM High IOPS PCIe

Adapters using Microsoft

Fast Track Data Warehouse for SQL Server 2012

Standard Configuration

Michael Lawson

Vinay Kulkarni

IBM Systems and Technology Group

November 18, 2013

Configuration guide for IBM Flex System x240 and IBM High IOPS PCIe Adapters using Microsoft Fast Track Data Warehouse for SQL Server 2012

2 Copyright IBM Corporation 2013

Table of Contents Table of Contents .................................................................................................................. 2

IBM Solutions with Microsoft Fast Track Data Warehouse for SQL Server 2012 ...................... 3

IBM Flex System servers 3

IBM High IOPS MLC Adapters 4

Building the Fast Track Data Warehouse (FTDW) Environment ............................................... 6

Configure OS and server power options 6

Install the High IOPS Adapters 6

Configure the adapter storage 6

Assign mount points to the storage partitions 7

Disk performance validation 7

Installing SQL Server 2012 7

SQL Server 2012 Tuning Options 7

Lock pages in memory 10

Creating a Fast Track Database 13

Spreading TempDB across available logical drives 13

Loading a Fast Track Database 14

Conclusion ........................................................................................................................... 14

For more information .......................................................................................................... 15

Appendix A OS and server power options ......................................................................... 15

Appendix B Split using Virtual Controller technology ......................................................... 17

Appendix C Create mount points and format volumes ....................................................... 17

Appendix D Install SQL Server ........................................................................................... 21

Appendix E Database creation code .................................................................................. 23

Appendix F - Tempdb relocation code .................................................................................. 26

Appendix G - SQLIO tests ..................................................................................................... 27

Trademarks and special notices ........................................................................................... 29

November 18, 2013



IBM Solutions with Microsoft Fast Track Data Warehouse for SQL Server

2012 The configuration guide for IBM Flex System servers using Microsoft Fast Track Data Warehouse for

SQL Server 2012 is designed to provide a methodology for creating a minimal and balanced architecture

for predictable data warehouse workloads.

In a Fast Track configuration, minimal server and storage hardware is purchased to match the processing

capability of the hardware to the data warehouse processing requirements of the organization. The goal

is to create a balanced configuration across all hardware and software elements. To accomplish this,

each hardware component is analyzed to determine the maximum data processing capabilities and then

matched to provide a balanced configuration. The configuration is then carefully tuned to achieve

maximum performance. This holistic approach minimizes the risk of purchasing hardware that will never

be fully utilized by SQL Server for data processing, yet provides ample room for capacity growth.

Accelerate your data warehouse roadmap with hardware configurations ready to support SQL Server 2012.

Virtually eliminate hardware testing and reduce tuning with high performance out of the box.

Scale from 5 to 112 terabytes using compression capabilities in SQL Server 2012 Enterprise.

Enjoy a lower total cost of ownership (TCO) through better price performance, rapid deployment, and industry-standard hardware.

IBM Flex System servers help provide the right performance for your needs and deliver enhanced, energy-efficient performance for data-intensive applications.

Configurations jointly reviewed by IBM and Microsoft for optimal data warehouse performance.

This document explains the various optimizations at both the hardware and software levels to provide a

complete reference architecture for achieving a baseline level of performance using IBM Flex System

x240 server and the IBM High IOPS MLC Adapters.

IBM Flex System servers IBM has published several Fast Track Data Warehouse for SQL Server 2012 configurations: a Basic one

on the System x3550 M4 server, a Standard one on the x3650 M4 server, an advanced one on the x3850

X5 server, and a High Availability one on the x3650 M4 server. This guide is based on the Flex System

x240 server and the High IOPS MLC Adapters. The Flex System x240 server is based on the Intel Xeon

processor E5-2600 series. The x240 is the right choice when compromise is not an option; it blends

outstanding uptime, performance and I/O flexibility for cost efficiency and rock-solid reliability. This

compact rack server offers an energy-smart, affordable and easy-to-use solution with a pay-as-you-grow

design to help lower costs and manage risks. With more computing power per watt, support for the

latest Intel Xeon E5-2600 series processors and advanced memory support, the x240 offers balanced

performance and density.

November 18, 2013



IBM Reference Configurations for Microsoft Fast Track Data Warehouse

http://www-03.ibm.com/systems/x/os/windows/fasttrack.html

Get more information about the IBM Flex System x240 at:

http://www-03.ibm.com/systems/flex/compute-node/x86/x240/

Get more information about the IBM System x3550 M4 at:

http://www-03.ibm.com/systems/x/hardware/rack/x3550m4/index.html

Get more information about the IBM System x3650 M4 at:

http://www-03.ibm.com/systems/x/hardware/rack/x3650m4/index.html

Get more information about the IBM System x3850 X5 at:

http://www-03.ibm.com/systems/x/hardware/enterprise/x3850x5/index.html

IBM High IOPS MLC Adapters The next generation of IBM High IOPS solid state PCIe adapters bring improved performance, scalability and increased capacity while maintaining industry leading endurance. The efficient modular design ensures low firmware impact, low software impact, and low route complexity. Additionally, the firmware of the next generation PCIe adapter has a new controller chip, 49-bit ECC, 75 watt slot support and Adaptive Flashback which provides protection from multiple NAND failures, infant mortality and premature wear out. Combined High IOPS adapters deliver industry leading low latency and hundreds of thousands of IOPS in a single server to power data warehouses and other big data applications.

Figure 1 IBM High IOPS MLC Adapter

Get more information about the IBM High IOPS Adapters at: http://www-03.ibm.com/systems/x/options/storage/solidstate/adapters.html

Important notice

This paper is designed to provide a methodology for sequential data workloads. Use of this methodology

to design an environment for non-sequential workloads might result in poor performance.

November 18, 2013



The configurations described in this paper have been tested by IBM engineers on the IBM servers and

storage. IBM makes no claims for performance on systems and storage other than those described in

this paper.

November 18, 2013



Building the Fast Track Data Warehouse (FTDW) Environment Microsoft Windows Server 2012 is the recommended Microsoft operating system for a Fast Track Data

Warehouse environment. This paper assumes that you are using Windows Server 2012. This solution is

built on one Flex System x240 server with four IBM 1.2TB High IOPS MLC Mono Adapters in an IBM Flex

System PCIe Expansion Node. The full hardware details are available here:

http://www.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/WP102366

Once you have installed Windows Server 2012, follow the steps outlined in this document to build the

Fast Track Data Warehouse environment. We recommend performing the steps in the following order:

1. Configure OS and server power options

2. Install the High IOPS Adapters.

3. Configure the adapter storage.

4. Assign mount points to the storage partitions.

5. Validate disk performance with SQLIO.

6. Install Microsoft SQL Server 2012.

7. Tune Microsoft SQL Server for sequential data access.

Configure OS and server power options The Windows and BIOS (uEFI) server power options should be configured for high performance. This will

allow the High IOPS Adapters to perform at their maximum performance. Refer to Appendix A OS

and server power options.

Install the High IOPS Adapters Attach the IBM Flex System PCIe Expansion Node to the Flex System x240 server. Install the four IBM

1.2TB High IOPS MLC Mono Adapters in the IBM Flex System PCIe Expansion Node, two in the full height

riser and two in the low height riser. You may need to replace the brackets on the adapters, if they are

not the correct height.

IBM Flex System x240 Compute Node Installation and Service Guide

http://publib.boulder.ibm.com/infocenter/flexsys/information/topic/com.ibm.acc.8737.doc/printable_doc.html

Installing the PCIe Expansion Node

http://pic.dhe.ibm.com/infocenter/flexsys/information/index.jsp?topic=%2Fcom.ibm.acc.8737.doc%2Fprintable_d

oc.html

Configure the adapter storage Obtain the latest IBM High IOPS Driver and Utilities for Microsoft Windows package from:

http://www-947.ibm.com/support/entry/portal/docdisplay?lndocid=MIGR-5091576

November 18, 2013



IBM High IOPS System x server configuration information and requirements http://www-947.ibm.com/support/entry/portal/docdisplay?lndocid=serv-iops

Install the driver. Use the utility to create 2 two Virtual Controllers on each device. Follow the

instructions in Appendix B Split using Virtual Controller technology.

Assign mount points to the storage partitions The previous step created the LUNs and presented them to the Fast Track server. The next step is to run

the scripts provided in Appendix C Create mount points and format volumes to create the mount

points for the volumes and to format and mount the logical drives in the operating system.

Disk performance validation After the storage volumes have been attached and mounted on the server, and before installing SQL

Server, it is a good idea to validate the performance of your disk volumes. This can be done with a tool

from Microsoft called SQLIO. This tool does not require SQL Server to be installed and it does not

attempt to simulate SQL Server I/O patterns (SQLIOSim, another tool, does that). SQLIO is a simple

synthetic disk I/O generation tool, which is similar to IOMeter. It is recommended to resolve any

performance issues identified before proceeding to install SQL Server. Refer to Appendix G - SQLIO

tests for the details of how to run the tests and the expected results.

Installing SQL Server 2012 SQL Server 2012 can be installed from the graphical user interface (GUI) or from the command line.

Refer to Appendix D Install SQL Server for the command-line install script.

SQL Server 2012 Tuning Options A typical Data Warehouse workload follows an I/O pattern of mostly large sequential reads. To optimize

SQL Server for a large sequential read workload, configure SQL server to start with the E option and to

use the 1117 trace flag by setting the T1117 parameter upon startup.

E

The change made to SQL Server by adding it as a startup parameter is that the round robin

behavior in the proportional fill algorithm is affected. The algorithm does not switch files until

64 extents have been allocated from a single file. Without this parameter, the algorithm

switches files after 1 extent has been allocated. This keeps more rows physically contiguous,

which speeds up table scans.

-T1117

This trace flag ensures equal expansion of all files within a file group.

These startup options can be configured using the SQL Server Configuration Manager program.

November 18, 2013



To configure the startup options, follow these steps:

1. Start the SQL Server Configuration Manager program by clicking Start All Programs

Microsoft SQL Server 2012 Configuration Tools SQL Server Configuration Manager.

2. Once SQL Server Configuration Manager loads, click on SQL Server Services in the left pane and

then right-click SQL Server in the right pane and select Properties.

November 18, 2013



3. Click on the Startup Parameters tab.

November 18, 2013



4. Add -E and -T1117 to the startup parameters one at a time.

5. Click OK.

6. Restart SQL Service for the new parameters to take effect.

7. Close the SQL Server Configuration Manager application.

Lock pages in memory This optimization prevents SQL Server buffer pages from being paged out by the operating system. This

improves performance when SQL Server is the only application running on the server. To enable this

optimization, provide the Lock Pages in Memory permission to the account that runs SQL Server.

November 18, 2013



1. Begin by opening the Local Group Policy Editor.

2. Navigate to Local Computer Policy Computer Configuration Windows Settings Security

Settings Local Policies User Rights Assignment and in the right pane, double-click Lock

pages in memory.

November 18, 2013



3. Add the user account under which SQL Server runs. Click OK.

November 18, 2013



4. Verify that the correct account has been granted permission for Lock pages in memory.

Creating a Fast Track Database When creating a Fast Track database, it is important to pre-allocate the space and to spread the

database across the available logical drives before loading the production data into the database.

The example script in Appendix E Database creation code creates a 512 GB database spread across

the logical drives available to the Fast Track server.

It is important to pre-allocate enough space before loading production data into the database to

guarantee that the database does not expand during data load. If this occurs, the data becomes

fragmented, affecting retrieval performance.

Spreading TempDB across available logical drives The Microsoft Fast Track documentation recommends spreading TempDB across available logical drives.

The example script in Appendix F - Tempdb relocation code demonstrates how to split TempDB across

the logical drives available to the Fast Track server. This improves TempDB performance by bringing

more disk drives into play and reducing contention between threads running on different processors.

November 18, 2013



Loading a Fast Track Database To obtain the full benefits of the Fast Track methodology, it is essential to follow the best practices for

loading the database. This is because the typical queries generate large block sequential reads, which

perform best on data that has been loaded preserving a high degree of physical sequential order of the

data rows. This complex topic depends on each customers unique database. The topic is addressed in

detail in the following documents:

Fast Track Data Warehouse Reference Guide for SQL Server 2012

http://download.microsoft.com/download/D/2/0/D20E1C5F-72EA-4505-9F26-

FEF9550EFD44/Fast%20Track%20DW%20Reference%20Guide%20for%20SQL%202012.docx?WT.mc_id=

aff-n-in-loc--pd

Fast Track Data Load Best Practices Overview

http://download.microsoft.com/download/2/F/6/2F6AF505-30B7-4505-80E5-

B363714441C7/FT_3.0_Data_Load_Best_Practices_Overview_2_14_11.pptx

Conclusion The IBM reference configurations for Fast Track Data Warehouse for SQL Server 2012 bring together the

right mix of technology and software, integrating the powerful IBM Flex System servers, robust IBM

System Storage, and the data warehouse capabilities of Microsoft SQL Server 2012 Enterprise Edition.

By using pretested and validated IBM configurations (specifically balanced and optimized for data

warehousing), the result is a fully tuned and documented solution that is ready to deploy. Fast Track

Data Warehouse for SQL Server 2012 for IBM Flex System are methodically tested and tuned to save you

months of configuration, setup, testing, and tuning. The combination results in a robust enterprise

solution built on a proven platform IBM Flex System servers and Microsoft SQL Server 2012.

November 18, 2013



For more information The following websites provide useful references to supplement the information contained in this paper.

IBM System x Servers:

ibm.com/systems/x

IBM Standalone Solutions Configuration Tool:

ibm.com/servers/eserver/xseries/library/configtools.html

IBM Configuration and Options Guide:

ibm.com/servers/eserver/xseries/cog

IBM ServerProven Program:

ibm.com/systems/info/x86servers/serverproven/compat/us/

IBM Technical Support:

ibm.com/server/support

Other IBM Technical Support Resources:

ibm.com/servers/eserver/techsupport.html

Appendix A OS and server power options In a PowerShell or command prompt issue these commands to set the Windows Power Scheme to High

Performance:

powercfg -l

$High_performance = "8c5e7fda-e8bf-4a96-9a85-a6e23a8c635c"

powercfg -S $High_performance

powercfg -l

Using the ASU tool, set the uEFI power settings as follows:

asu64.exe set Processors.C1EnhancedMode Disable

asu64.exe set Processors.ProcessorPerformanceStates Disable

Then restart the server.

The ASU tool can be downloaded here:

IBM Advanced Settings Utility (ASU)

November 18, 2013



https://www-947.ibm.com/support/entry/myportal/docdisplay?lndocid=TOOL-ASU

November 18, 2013



Appendix B Split using Virtual Controller technology Refer to the IBM ioMemory VSL x.x.x User Guide for Microsoft Windows pdf included in the IBM High

IOPS Driver and Utilities for Microsoft Windows package for instructions on Virtual Controller

Conversion. Here are sample commands for splitting the controllers:

fio-detach /dev/fct

fio-detach /dev/fct

fio-detach /dev/fct

fio-detach /dev/fct

fio-update-iodrive --split ""

fio-attach /dev/fct

fio-attach /dev/fct

fio-attach /dev/fct

fio-attach /dev/fct

fio-attach /dev/fct

fio-attach /dev/fct

fio-attach /dev/fct

fio-attach /dev/fct

These commands detach the High IOPS devices, split all the devices into 2 devices (and update the

firmware), and attach the devices. Run the detach and attach commands once for each device. That

will be 4 detach commands and 8 attach commands. The fio-update-iodrive command updates

all the devices at once, so it only needs to be run one time. When this step is completed, there should

be 8 devices (LUNs) visible in Windows (in Disk Manager, for example).

Appendix C Create mount points and format volumes This section includes the commands to create the mount points for the volumes, and to format and

mount the logical drives in the operating system.

Use the following commands on the Windows server to create mount point directories:

mkdir C:\FT\PRI\SE1-SP1-DG1-v1







November 18, 2013



mkdir C:\FT\LOG\SE4-SP2-DG8-v8

After the logical drives have been presented to the Fast Track server, the following PowerShell script

can be used to bring them online, place them in mount points, and format them. You must first

determine the Windows disk number of each logical drive, which you can do using the Disk Manager

snap-in. Right-click the disk (on the left side), click Properties and you can see the slot number and the

serial number. Use the naming convention to identify the slot in the SE part of the name. Use

the SP1 to name the first virtual controller and SP2 to name the second virtual controller (for the second

one, the serial number ends with P1) in the same slot. Change the to the corresponding Windows

disk number for each logical drive.

EXAMPLE 1

Suppose Disk 0 is in slot 2 and has a serial number without the suffix P1. Then it should have the

name SE2-SP1-DG3-v3. Locate the portion of the PowerShell script for v3 and change

$Disk_Num =

To

$Disk_Num = 0

EXAMPLE 2

Suppose Disk 7 is in slot 1 and has a serial number with the suffix P1. Then it should have the name

SE1-SP2-DG2-v2. Locate the portion of the PowerShell script for v2 and change

$Disk_Num =

To

$Disk_Num = 7

Using this naming convention allows you to identify each adapter with each volume. It also is assumed

by the SQLIO scripts.

$TopDir = "C:\FT\PRI\"

# v1

$Disk_Num =

$SubDir = "SE1-SP1-DG1-v1"

$FullDir = $TopDir + $SubDir + "\"

Clear-Disk -Number $Disk_Num -RemoveData -RemoveOEM -Confirm:$false

Set-Disk -Number $Disk_Num -IsOffline $false

Set-Disk -Number $Disk_Num -isReadOnly $false

November 18, 2013



Initialize-Disk -Number $Disk_Num -PartitionStyle GPT

New-Partition -Disknumber $Disk_Num -UseMaximumSize `

| Format-Volume -FileSystem NTFS -AllocationUnitSize 65536 `

- NewFileSystemLabel $SubDir -Confirm:$false

Add-PartitionAccessPath -Disknumber $Disk_Num -Partition 2 `

-AccessPath $FullDir

# v2

$Disk_Num =












# v3

$Disk_Num =












# v4

$Disk_Num =







November 18, 2013








# v5

$Disk_Num =












# v6

$Disk_Num =












# v7

$Disk_Num =








November 18, 2013







# v8

$TopDir = "C:\FT\LOG\"

$Disk_Num =












Appendix D Install SQL Server To install SQL Server from the command line, create the parameter file

(SQL_Engine_ConfigurationFile.ini) given below, with your changes and execute the following commands

(put them in SQL_Engine_Install.cmd), with the SQL install media (for example, a mounted ISO) in drive

E:\. Customize this with the path to the configuration file, the password for the SQLServiceEngineUser

account, the password for the SQLServiceAgentUser account, and the SQL Server Product ID. Note the

text beginning with Setup.exe and ending with IACCEPTSQLSERVERLICENSETERMS is all one line.

SQL_Engine_Install.cmd

pushd

cd /D "E:\"

Setup.exe /ConfigurationFile="\SQL_Engine_ConfigurationFile.ini"

/SQLSVCPASSWORD=" /AGTSVCPASSWORD=""

/PID="" /IACCEPTSQLSERVERLICENSETERMS

popd

Create the following set of parameters in a file named SQL_Engine_ConfigurationFile.ini. Customize this

with your domain and SQLAdminUser, SQLServiceEngineUser and SQLServiceAgentUser account names.

This will install the SQL Server Engine component and SQL Server Management Studio.

November 18, 2013



SQL_Engine_ConfigurationFile.ini

;SQL Server 2012 Configuration File

[OPTIONS]

ACTION="Install"

ENU="True"

QUIET="False"

QUIETSIMPLE="True"

UpdateEnabled="True"

FEATURES=SQLENGINE,SSMS,ADV_SSMS

UpdateSource="MU"

HELP="False"

INDICATEPROGRESS="True"

X86="False"

INSTALLSHAREDDIR="C:\Program Files\Microsoft SQL Server"

INSTALLSHAREDWOWDIR="C:\Program Files (x86)\Microsoft SQL Server"

INSTANCENAME="MSSQLSERVER"

SQMREPORTING="True"

INSTANCEID="MSSQLSERVER"

ERRORREPORTING="True"

INSTANCEDIR="C:\Program Files\Microsoft SQL Server"

AGTSVCACCOUNT="\"

AGTSVCSTARTUPTYPE="Manual"

COMMFABRICPORT="0"

COMMFABRICNETWORKLEVEL="0"

COMMFABRICENCRYPTION="0"

MATRIXCMBRICKCOMMPORT="0"

SQLSVCSTARTUPTYPE="Automatic"

FILESTREAMLEVEL="0"

ENABLERANU="False"

SQLCOLLATION="SQL_Latin1_General_CP1_CI_AS"

SQLSVCACCOUNT="\"

SQLSYSADMINACCOUNTS="\"

ADDCURRENTUSERASSQLADMIN="False"

TCPENABLED="1"

NPENABLED="0"

BROWSERSVCSTARTUPTYPE="Automatic"

This SQL Server install requires Microsoft .NET Framework 3.5. If you need to install it, mount the

Windows Server 2012 installation media (on F: for example) and run this command (all on one line) from

an elevated command or PowerShell prompt, before installing SQL Server:

November 18, 2013



dism /online /enable-feature /featurename:netfx3 /all

/source:F:\sources\sxs

Appendix E Database creation code The following is sample T-SQL code to create a Fast Track database on the storage as configured above.

This database has file groups for non-partitioned data, staging data, partitioned data, and a log.

CREATE DATABASE [FT_Demo] ON PRIMARY /*****Small primary file on primary filegroup ****************************/ (NAME=N'FT_Demo',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo.mdf',SIZE=204800KB, MAXSIZE=UNLIMITED, FILEGROWTH=0), /***************************************************************** Normally each file would be created on a different LUN through each mount points *****************************************************************/ FILEGROUP [Base] DEFAULT (NAME=N'FT_Demo_Base_1',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_Base_1.ndf',SIZE=14629MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_Base_2',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_Base_2.ndf',SIZE=14629MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_Base_3',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_Base_3.ndf',SIZE=14629MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_Base_4',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_Base_4.ndf',SIZE=14629MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_Base_5',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_Base_5.ndf',SIZE=14629MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_Base_6',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_Base_6.ndf',SIZE=14629MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_Base_7',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_Base_7.ndf',SIZE=14629MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), FILEGROUP [stage_part_ci] (NAME=N'FT_Demo_stage_part_ci_1',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_stage_part_ci_1.ndf',SIZE=46812MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_stage_part_ci_2',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_stage_part_ci_2.ndf',SIZE=46812MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_stage_part_ci_3',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_stage_part_ci_3.ndf',SIZE=46812MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_stage_part_ci_4',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_stage_part_ci_4.ndf',SIZE=46812MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_stage_part_ci_5',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_stage_part_ci_5.ndf',SIZE=46812MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_stage_part_ci_6',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_stage_part_ci_6.ndf',SIZE=46812MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_stage_part_ci_7',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_stage_part_ci_7.ndf',SIZE=46812MB,FILEGROWTH=0,MAXSIZE=UNLIMITED),

November 18, 2013



FILEGROUP [part_ci1FG] (NAME=N'FT_Demo_part_ci1_01',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_part_ci1_01.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci1_02',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_part_ci1_02.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci1_03',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_part_ci1_03.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci1_04',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_part_ci1_04.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci1_05',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_part_ci1_05.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci1_06',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_part_ci1_06.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci1_07',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_part_ci1_07.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), FILEGROUP [part_ci2FG] (NAME=N'FT_Demo_part_ci2_01',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_part_ci2_01.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci2_02',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_part_ci2_02.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci2_03',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_part_ci2_03.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci2_04',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_part_ci2_04.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci2_05',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_part_ci2_05.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci2_06',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_part_ci2_06.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci2_07',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_part_ci2_07.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), FILEGROUP [part_ci3FG] (NAME=N'FT_Demo_part_ci3_01',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_part_ci3_01.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci3_02',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_part_ci3_02.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci3_03',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_part_ci3_03.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci3_04',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_part_ci3_04.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci3_05',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_part_ci3_05.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci3_06',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_part_ci3_06.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci3_07',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_part_ci3_07.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), FILEGROUP [part_ci4FG] (NAME=N'FT_Demo_part_ci4_01',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_part_ci4_01.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci4_02',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_part_ci4_02.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci4_03',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_part_ci4_03.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci4_04',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_part_ci4_04.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED),

November 18, 2013



(NAME=N'FT_Demo_part_ci4_05',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_part_ci4_05.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci4_06',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_part_ci4_06.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci4_07',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_part_ci4_07.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), FILEGROUP [part_ci5FG] (NAME=N'FT_Demo_part_ci5_01',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_part_ci5_01.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci5_02',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_part_ci5_02.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci5_03',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_part_ci5_03.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci5_04',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_part_ci5_04.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci5_05',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_part_ci5_05.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci5_06',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_part_ci5_06.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci5_07',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_part_ci5_07.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), FILEGROUP [part_ci6FG] (NAME=N'FT_Demo_part_ci6_01',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_part_ci6_01.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci6_02',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_part_ci6_02.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci6_03',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_part_ci6_03.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci6_04',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_part_ci6_04.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci6_05',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_part_ci6_05.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci6_06',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_part_ci6_06.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci6_07',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_part_ci6_07.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), FILEGROUP [part_ci7FG] (NAME=N'FT_Demo_part_ci7_01',FILENAME=N'c:\ft\pri\se1-sp1-dg1-v1\FT_Demo_part_ci7_01.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci7_02',FILENAME=N'c:\ft\pri\se1-sp2-dg2-v2\FT_Demo_part_ci7_02.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci7_03',FILENAME=N'c:\ft\pri\se2-sp1-dg3-v3\FT_Demo_part_ci7_03.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci7_04',FILENAME=N'c:\ft\pri\se2-sp2-dg4-v4\FT_Demo_part_ci7_04.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci7_05',FILENAME=N'c:\ft\pri\se3-sp1-dg5-v5\FT_Demo_part_ci7_05.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci7_06',FILENAME=N'c:\ft\pri\se3-sp2-dg6-v6\FT_Demo_part_ci7_06.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED), (NAME=N'FT_Demo_part_ci7_07',FILENAME=N'c:\ft\pri\se4-sp1-dg7-v7\FT_Demo_part_ci7_07.ndf',SIZE=20898MB,FILEGROWTH=0,MAXSIZE=UNLIMITED) /****************************************************************

November 18, 2013



Log will go on LUN created for logs 1 LUN per disk enclosure is created for logs ****************************************************************/ LOG ON (NAME=N'FT_Demo_LOG',FILENAME = N'C:\FT\LOG\SE4-SP2-DG8-v8\FT_Demo_log.LDF',SIZE=50GB,MAXSIZE=50GB,FILEGROWTH=10%) GO /*****************Configure recommended settings***********************/ ALTER DATABASE FT_Demo SET AUTO_CREATE_STATISTICS ON GO ALTER DATABASE FT_Demo SET AUTO_UPDATE_STATISTICS ON GO ALTER DATABASE FT_Demo SET RECOVERY SIMPLE GO

Appendix F - Tempdb relocation code Sample T-SQL code to spread the Tempdb database across the storage as configured above. This code

puts one file on each logical disk. Run this code in two pieces, restarting SQL where indicated in the

comments.

USE [master] GO -- MOVE FILES from C:\ ALTER DATABASE tempdb MODIFY FILE ( NAME = N'templog', FILENAME = N'C:\FT\LOG\SE4-SP2-DG8-v8\tempdb.ldf' , SIZE = 10MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10%) GO ALTER DATABASE tempdb MODIFY FILE ( NAME = N'tempdev', FILENAME = N'C:\Program Files\Microsoft SQL Server\MSSQL11.MSSQLSERVER\MSSQL\DATA\tempdb.mdf' , SIZE = 10MB , MAXSIZE = UNLIMITED, FILEGROWTH = 102400KB ) GO ALTER DATABASE tempdb MODIFY FILE ( NAME = N'tempdev', FILENAME = N'C:\FT\PRI\SE1-SP1-DG1-v1\tempdb.mdf' , SIZE = 10MB , MAXSIZE = UNLIMITED, FILEGROWTH = 102400KB ) GO -- RESTART SQL ALTER DATABASE tempdb MODIFY FILE ( NAME = N'templog', FILENAME = N'C:\FT\LOG\SE4-SP2-DG8-v8\tempdb.ldf' , SIZE = 50GB , MAXSIZE = UNLIMITED, FILEGROWTH = 10%)

November 18, 2013



GO ALTER DATABASE tempdb MODIFY FILE ( NAME = N'tempdev', FILENAME = N'C:\FT\PRI\SE1-SP1-DG1-v1\tempdb.mdf' , SIZE = 45GB , MAXSIZE = UNLIMITED, FILEGROWTH = 1GB ) GO -- ADD FILES - 1 on each LUN ALTER DATABASE tempdb ADD FILE ( NAME = N'tempdev02', FILENAME = N'C:\FT\PRI\SE1-SP2-DG2-v2\tempdev02.ndf' , SIZE = 45GB , MAXSIZE = UNLIMITED, FILEGROWTH = 1GB ) GO ALTER DATABASE tempdb ADD FILE ( NAME = N'tempdev03', FILENAME = N'C:\FT\PRI\SE2-SP1-DG3-v3\tempdev03.ndf' , SIZE = 45GB , MAXSIZE = UNLIMITED, FILEGROWTH = 1GB ) GO ALTER DATABASE tempdb ADD FILE ( NAME = N'tempdev04', FILENAME = N'C:\FT\PRI\SE2-SP2-DG4-v4\tempdev04.ndf' , SIZE = 45GB , MAXSIZE = UNLIMITED, FILEGROWTH = 1GB ) GO ALTER DATABASE tempdb ADD FILE ( NAME = N'tempdev05', FILENAME = N'C:\FT\PRI\SE3-SP1-DG5-v5\tempdev05.ndf' , SIZE = 45GB , MAXSIZE = UNLIMITED, FILEGROWTH = 1GB ) GO ALTER DATABASE tempdb ADD FILE ( NAME = N'tempdev06', FILENAME = N'C:\FT\PRI\SE3-SP2-DG6-v6\tempdev06.ndf' , SIZE = 45GB , MAXSIZE = UNLIMITED, FILEGROWTH = 1GB ) GO ALTER DATABASE tempdb ADD FILE ( NAME = N'tempdev07', FILENAME = N'C:\FT\PRI\SE4-SP1-DG7-v7\tempdev07.ndf' , SIZE = 45GB , MAXSIZE = UNLIMITED, FILEGROWTH = 1GB ) GO -- RESTART SQL SELECT * FROM sys.master_files WHERE database_id = DB_ID('tempdb'); GO

Appendix G - SQLIO tests You can download the SQLIO tool from the following URL.

http://www.microsoft.com/download/en/details.aspx?id=20163

The scripts for running the tests can be requested from the authors ([email protected],

[email protected]).

First create 7 SQLIO test files of 25 GB on each of the 7 data LUNs.

November 18, 2013



The parameters to SQLIO are always the same, except for varying the LUN. The disk access specification

is: 512 KB block sequential read using two threads with 60 outstanding I/Os on each thread, run for 60

seconds. Here is an example for a test using the first LUN:

set lun=C:\FT\PRI\SE1-SP1-DG1-v1\sqlio_25GB.dat

set rtime=60

set thr=2

set out=60

set blk=512

set ransq=sequential

set rw=R

sqlio.exe -f%ransq% -k%rw% -b%blk% -o%out% -t%thr% -s%rtime% -LS %lun%

Run the following tests one at a time. The following are the expected results for each test in MB per

second:

1. [1,258] Each LUN separately (seven tests: V1, V2, V3, V4, V5, V6, V7).

2. [1,432] Each High IOPS Adapter with two LUNs at the same time, separately (three tests: V1+V2,

V3+V4, V5+V6).

3. [5,548] All seven LUNs at the same time (one test: V1+V2+V3+V4+V5+V6+V7).

November 18, 2013



Trademarks and special notices Copyright IBM Corporation 2013

References in this document to IBM products or services do not imply that IBM intends to make them

available in every country.

IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business

Machines Corporation in the United States, other countries, or both. If these and other IBM

trademarked terms are marked on their first occurrence in this information with a trademark symbol (

or ), these symbols indicate U.S. registered or common law trademarks owned by IBM at the time this

information was published. Such trademarks may also be registered or common law trademarks in other

countries. A current list of IBM trademarks is available on the Web at "Copyright and trademark

information" at www.ibm.com/legal/copytrade.shtml.

Java and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or

its affiliates.

Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in

the United States, other countries, or both.

Intel, Intel Inside (logos), MMX, and Pentium are trademarks of Intel Corporation in the United States,

other countries, or both.

UNIX is a registered trademark of The Open Group in the United States and other countries.

Linux is a trademark of Linus Torvalds in the United States, other countries, or both.

SET and the SET Logo are trademarks owned by SET Secure Electronic Transaction LLC.

Other company, product, or service names may be trademarks or service marks of others.

Information is provided "AS IS" without warranty of any kind.

All customer examples described are presented as illustrations of how those customers have used IBM

products and the results they may have achieved. Actual environmental costs and performance

characteristics may vary by customer.

Information concerning non-IBM products was obtained from a supplier of these products, published

announcement material, or other publicly available sources and does not constitute an endorsement of

such products by IBM. Sources for non-IBM list prices and performance numbers are taken from publicly

available information, including vendor announcements and vendor worldwide homepages. IBM has not

tested these products and cannot confirm the accuracy of performance, capability, or any other claims

November 18, 2013



related to non-IBM products. Questions on the capability of non-IBM products should be addressed to

the supplier of those products.

All statements regarding IBM future direction and intent are subject to change or withdrawal without

notice, and represent goals and objectives only. Contact your local IBM office or IBM authorized reseller

for the full text of the specific Statement of Direction.

Some information addresses anticipated future capabilities. Such information is not intended as a

definitive statement of a commitment to specific levels of performance, function or delivery schedules

with respect to any future products. Such commitments are only made in IBM product announcements.

The information is presented here to communicate IBM's current investment and development activities

as a good faith effort to help with our customers' future planning.

Performance is based on measurements and projections using standard IBM benchmarks in a controlled

environment. The actual throughput or performance that any user will experience will vary depending

upon considerations such as the amount of multiprogramming in the user's job stream, the I/O

configuration, the storage configuration, and the workload processed. Therefore, no assurance can be

given that an individual user will achieve throughput or performance improvements equivalent to the

ratios stated here.

Photographs shown are of engineering prototypes. Changes may be incorporated in production models.

Any references in this information to non-IBM websites are provided for convenience only and do not in

any manner serve as an endorsement of those websites. The materials at those websites are not part of

the materials for this IBM product and use of those websites is at your own risk.

IBM Fast Track DataWarehouse SQLServer 2012 Std x240 7TB ConfigurationGuide

Documents

Transcript of IBM Fast Track DataWarehouse SQLServer 2012 Std x240 7TB ConfigurationGuide