Resource Manager Guide

Resource Manager Administrator’s Guide
For other guides in this document set, go to the Document Center
Resource Manager for MetaFrame® Presentation Server

Citrix® MetaFrame Presentation Server 4.0,
Enterprise Edition for Windows®
Citrix MetaFrame Access Suite
Use of the product documented in this guide is subject to your prior acceptance of the End User License
Agreement. Note that copies of the End User License Agreement are included in the root directory of the
MetaFrame Presentation Server CD-ROM and in the root directory of the Components CD-ROM.
Information in this document is subject to change without notice. Companies, names, and data used in
examples herein are fictitious unless otherwise noted. Other than printing one copy for personal use, no
part of this document may be reproduced or transmitted in any form or by any means, electronic or
mechanical, for any purpose, without the express written permission of Citrix Systems, Inc.
Copyright © 2003-2005 Citrix Systems, Inc. All rights reserved.
Citrix, ICA (Independent Computing Architecture), MetaFrame, MetaFrame XP, NFuse, and Program
Neighborhood are registered trademarks, and Citrix Solutions Network and SpeedScreen are trademarks
of Citrix Systems Inc. in the United States and other countries.
RSA Encryption © 1996-1997 RSA Security Inc., All Rights Reserved.
Adobe, Acrobat, and PostScript are trademarks or registered trademarks of Adobe Systems Incorporated
in the U.S. and/or other countries.
DB2 is a registered trademark and PowerPC is a trademark of International Business Machines Corp. in
the U.S. and other countries.
HP OpenView is a trademark of the Hewlett-Packard Company.
Microsoft, MS-DOS, Windows, Windows NT, Win32, Outlook, ActiveX, and Active Directory are either
registered trademarks or trademarks of Microsoft Corp. in the United States and/or other countries.
Tivoli and NetView are registered trademarks of International Business Machines Corp. in the U.S. and
other countries.
Unicenter is a registered trademark of Computer Associates International, Inc.
All other trademarks and registered trademarks are the property of their owners.
Last Updated: February 17, 2005 (SOC)
Go to Document Center Contents 3
Contents
Chapter 1 Introduction
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
About this Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Accessing Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Introducing Resource Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Monitoring Your Existing Server Farm . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Identifying, Diagnosing, and Solving Problems. . . . . . . . . . . . . . . . . . . . 11
Gauging and Justifying Future Resource Needs . . . . . . . . . . . . . . . . . . . 12
Planning and Scaling Your Server Farm . . . . . . . . . . . . . . . . . . . . . . . . . 12
Billing Users for Resource Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Delegating Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Receiving Failed Import Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Receiving License Server Connection Failure Alerts . . . . . . . . . . . . . . . 13
Resource Manager and the Access Suite Console. . . . . . . . . . . . . . . . . . . . . . . . . . 13
Extended Reporting Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Extended Real-Time Monitoring Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . 15
What to Read Next . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Chapter 2 Installing Resource Manager

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Software Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Licensing Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Installing Resource Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Before You Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Selecting a Server to Use as the Farm Metric Server. . . . . . . . . . . . . . . . 19
If You Are Upgrading from a Previous Version . . . . . . . . . . . . . . . . . . . . . . . . 19
Installing or Upgrading Resource Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Uninstalling Resource Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Setting Up a Summary Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Setting a System Data Source Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Configuring a Database Connection Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Turning the Summary Database On . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Turning the Summary Database Off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4 Resource Manager Administrator’s Guide Go to Document Center
Displaying Resource Manager and its Components . . . . . . . . . . . . . . . . . . . . . . . . 27

Displaying the Main Resource Manager Screen . . . . . . . . . . . . . . . . . . . . . . . . 28
Displaying Resource Manager for the Entire Server Farm . . . . . . . . . . . . . . . . 28
Displaying Resource Manager for a Single Server . . . . . . . . . . . . . . . . . . . . . . 29
Displaying the Applications in a Server Farm . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Changing the Location of Resource Manager after Installation . . . . . . . . . . . . . . . 30
Chapter 3 Monitoring Servers and Applications in Real Time

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Configuring Metrics on Servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Checking the Status of Server Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Customizing Server Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Running a Script on Transition to Red . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Monitoring the Disk on a Windows 2000 or 2003 Server . . . . . . . . . . . . 35
Configuring Application Metrics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Monitoring Your Farm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Suspending Notification of a Metric’s Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Preparing Your System for Resource Manager Alerts . . . . . . . . . . . . . . . . . . . . . . 37
Preparing Your System for Email Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Using MAPI to Send Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Setting up SSL for SMTP Email Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Preparing Your System for SMS Alerts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Configuring Resource Manager to Use SMS . . . . . . . . . . . . . . . . . . . . . . 41
Preparing Your System for SNMP Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Getting More Information About Metrics and Monitoring. . . . . . . . . . . . . . . . . . . 42
Chapter 4 Recording the History of Servers and Applications

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Resource Manager and the Summary Database . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
What Information Should I Record in the Database? . . . . . . . . . . . . . . . . . . . . . . . 45
Issues to Consider when Selecting Database Information. . . . . . . . . . . . . . . . . 45
The Benefits of Keeping Long-Term Resource Manager Information . . 45
Deciding which Metrics Information to Store . . . . . . . . . . . . . . . . . . . . . 45
How Can Recording a Metric Affect Database Growth?. . . . . . . . . . . . . 46
Selecting Server and Application Metrics to Record in the Database. . . . . . . . 46
Using Farm-Wide Options for Reducing Summary Data . . . . . . . . . . . . . . . . . 46
Scheduling Summary Data Collection and Removal . . . . . . . . . . . . . . . . . . . . . . . 47

Ignoring Server Metrics during Periods of Low Server Activity . . . . . . . . . . . 47
Removing Unwanted Information from the Database. . . . . . . . . . . . . . . . . . . . 47
Setting a Purging Schedule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Setting the Summary Database Update Time . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Monitoring the Status of the Summary Database . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Managing the Resource Manager Summary Database . . . . . . . . . . . . . . . . . . . . . . 52
Estimating Summary Database Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Managing Summary Database Growth . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Chapter 5 Reporting and Analyzing Resource Manager Information

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Generating Reports to Analyze Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Creating Reports on Current Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Reporting on the Use of Processes or Applications. . . . . . . . . . . . . . . . . . . . . . 57
Reporting on User Activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Looking Back to a Specific Time on a Server . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Creating Reports on Past Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Specific Processor Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Looking Back at Specific Users’ Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Activities on a Specific Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Displaying Reports from Servers in Different Time Zones or Locales . . . . . . . . . 67
Reporting Server Is in a Different Time Zone . . . . . . . . . . . . . . . . . . . . . . . . . . 67
What if a Reporting Server Uses a Different Language?. . . . . . . . . . . . . . . . . . 68
Estimating the Concurrent User Capacity of a Server. . . . . . . . . . . . . . . . . . . . . . . 68
Chapter 6 Billing Users for Resource Usage

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Creating a Fee Profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Organizing Users into Cost Centers for Billing. . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Producing Billing Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Cost Center Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Domain User Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Information Presentation in the Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Chapter 7 Troubleshooting
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Unexpected Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Locating Servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Permissions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Appendix A Default Metric Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Default Set of Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Data Store Connection Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Minutes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Logical Disk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
% Disk Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
% Free Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Available Bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Pages/sec. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Network Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Bytes Total/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Paging File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
% Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
% Interrupt Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
% Processor Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Context Switches/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Terminal Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Active Sessions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Inactive Sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Appendix B Summary Database Schema . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Database Entity Relationship Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Database Entity Table Descriptions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Application History. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
SDB_APPHISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Application Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
SDB_APPMETRICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Client History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
SDB_CLIENTHISTORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Connection History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
SDB_CONNECTIONHISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Event Log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
SDB_EVENTLOG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Administrator Configurable Server Metrics. . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
SDB_METRICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
SDB_PROCESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
User Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
SDB_SESSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Version Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
SCHEMAVERSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Support and Look-Up Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
LU_APPNAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
LU_CLIENT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
LU_CLIENTPROPERTIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
LU_CLIENTTYPEMAPPINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
LU_FARMNAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
LU_INSTANCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
LU_LAUNCHER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
LU_METRIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
LU_METRICCOUNTER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
LU_NETDOMAIN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
LU_OBJECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
LU_PATH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
LU_PROCESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
LU_PROCESSNAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
LU_SERVER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
LU_SERVERNAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
LU_SERVERINF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
LU_WINSTATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
SDB_SCRATCH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
SDB_HEURISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
SQL Data Type Mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Go to Document Center
CHAPTER 1
Introduction
Overview
Welcome to Resource Manager for MetaFrame Presentation Server. This chapter
introduces you to the documentation and to Resource Manager. Topics include:
• How to use this guide
• Accessing documentation in general
• An introduction to Resource Manager
• A description of some extended Resource Manager features in the Access Suite
Console
• Finding more information about Resource Manager
About this Guide

This guide is for MetaFrame Administrators responsible for installing, configuring,
and maintaining Resource Manager.
This guide is designed to introduce you to the key concepts of Resource Manager,
and to get you started quickly and easily. This chapter introduces the documentation
and the Resource Manager product, and describes some extended Resource
Manager features available from the Access Suite Console. Subsequent chapters
provide an overview of how to deploy and configure Resource Manager, and
provide you with a summary of the main ways you can use it in your deployment.
This guide assumes knowledge of MetaFrame Presentation Server. If you will be
recording the history of server and application metrics using the summary database,
you need to be familiar with Microsoft SQL Server or Oracle DBMS.
For detailed information on how to carry out the Resource Manager procedures
outlined in this guide, see the Resource Manager help system.
Accessing Documentation
This administrator’s guide is part of the MetaFrame Presentation Server
documentation set. The documentation set includes online guides that correspond to
different features of MetaFrame Presentation Server. Online documentation is
provided as Adobe Portable Document Format (PDF) files.
Use the Document Center to access the complete set of online guides. The
Document Center provides a single point of access to the documentation that
enables you to go straight to the section that you need. The Document Center
includes:
• A list of common tasks and a link to each item of documentation.
• A search function that covers all the PDF guides. This is useful when you need
to consult a number of different guides.
• Cross-references between documents. You can move among documents as often
as you need using the links to other guides and the links to the Document
Center.
Important To view, search, and print the PDF documentation, you need to have the
Adobe Reader 5.0.5 or a later version with Search. You can download Adobe
Reader for free from Adobe Systems’ Web site at
http://www.adobe.com/.
If you prefer to access the guides without using the Document Center, you can
navigate to the component PDF files using Windows Explorer. If you prefer to use
printed documentation, you can also print each guide from Adobe Reader.
More information about Citrix documentation, and details about how to obtain
further information and support, is included in Getting Started with MetaFrame
Presentation Server.
Go to Document Center Chapter 1 Introduction 11
Introducing Resource Manager

You can use Resource Manager to manage resources on single or multiple
computers running MetaFrame Presentation Server. Resource Manager enables you
to collect, display, store, and analyze data about system performance, application or
process use, and user activity.
Use Resource Manager to:
• Watch what is happening at a particular moment for a system. This is known as
real-time monitoring.
• Analyze and report using records of system activity. You can create reports
about current activities or past activities using a summary database.
• Create Billing reports to charge users for their use of resources using a summary
database.
Resource Manager can track and store information about a wide variety of system
and network processes and events. These are known as metrics. If the value of a
metric falls outside normal limits, Resource Manager can inform you. During
installation, Resource Manager automatically creates a set of default metrics and
assigns limits to define the normal operation of each one.
Tip Resource Manager can track any Windows Performance Monitor counter as a
server metric. Further information on these counters is available in Appendix C of
the MetaFrame Presentation Server Administrator’s Guide and the Microsoft
Windows Help for Performance Monitor.
You can use Resource Manager in your server farm to help you with the following
tasks:
Monitoring Your Existing Server Farm

You can use Resource Manager to monitor and analyze system performance,
loading, and user behavior.
You can view information about an entire server farm, analyze individual servers
and applications, or monitor specific aspects of performance.
By fine-tuning the monitoring process, you can customize the information that
Resource Manager provides to suit your specific environment.
Identifying, Diagnosing, and Solving Problems

Resource Manager can warn you about any developing problems in your
environment. If a problem does occur, you can analyze the relevant data to help you
decide what action to take.
Gauging and Justifying Future Resource Needs

You can produce reports about system usage that help you identify requirements for
future resources, such as new servers or additional licenses.
Planning and Scaling Your Server Farm

By deploying Resource Manager on a pilot system, you can gauge how the server
farm will perform in different possible scenarios. This will help you to scale your
network, set baselines, and spot potential problems before they affect the final
installation. For example, you can evaluate whether or not a particular server can
support the activity of the desired number of users.
Billing Users for Resource Usage

You can produce Billing reports that use the resource usage data from the summary
database and a fee profile to calculate the charges for users of the server farm. You
define fee profiles to reflect different charging rates and currencies. When Resource
Manager generates a Billing report, it calculates the charges by multiplying the
resources used during the report period by the associated fees.
Delegating Administration
Enhanced delegated administration in Resource Manager allows you to delegate the
tasks of controlling applications and analyzing system and application data to
custom administrators. For more information on Enhanced Delegated
Administration, see the MetaFrame Presentation Server Administrator’s Guide.
To simplify the configuration of custom administrators, full administrators may find
it easier to first create a suitable folder hierarchy and then create custom
administrators with the necessary permissions. You create custom administrators
using the Add MetaFrame Administrator wizard in the Presentation Server
Console.
For Resource Manager you can:
• Delegate the control of Resource Manager applications to custom
administrators. A Resource Manager application is not published by
MetaFrame Presentation Server but is still recognized by the Resource
Manager system. Details of how to set up Resource Manager applications are
provided in the Resource Manager online help.
• Determine which custom administrators can generate current reports, summary
reports, or Billing reports.
• Ensure that administrators receive alerts only for those servers for which they
have responsibility.
Receiving Failed Import Alerts

Resource Manager warns you of failed attempts to update the summary database,
and can be configured to send alerts under these circumstances. In its red state, an
icon on the Summary Database tab identifies failures to commit any record in the
last attempted update to the database, and the Summary Database Configuration
dialog box allows you to notify administrators of any such failure. For further
information, see “Monitoring the Status of the Summary Database” on page 49 and
“Setting Up a Summary Database” on page 21.
Receiving License Server Connection Failure Alerts

With the Citrix MetaFrame Access Suite Licensing feature, Citrix licensing is
handled by one or more license servers. Computers running MetaFrame
Presentation Server communicate with license servers to ensure that client sessions
are licensed appropriately. A license server connection failure alert is raised when a
server cannot communicate with its associated license server. This may be due to a
hardware failure on the license server, the license service on the license server
malfunctioning, or network problems between the license server and the computer
running MetaFrame Presentation Server.
When contact with a license server is lost, the computer running MetaFrame
Presentation Server lapses into a licensing grace period. Typically, the licensing
grace period is 30 days. During this period, the MetaFrame Presentation Server
software is fully functional and connections to the server work normally. Contact
with the license server must be reestablished before the grace period ends, or the
software is reduced automatically to single user mode and only the administrator
can log on to the server.
For further information, see “Preparing Your System for Resource Manager Alerts”
on page 37.
Resource Manager and the Access Suite Console

Extended Reporting Capabilities
In addition to the reporting and monitoring capabilities provided by Resource
Manager in the Presentation Server Console, you can also create reports and
analyze your farms’ performance using two extensions to the Access Suite Console:
Report Center and Dashboard.
The Report Center is one of the system management tools provided through the
Access Suite Console. The Access Suite Console snaps into the Microsoft
Management Console to provide a central location for system management tools.
You can also use the console to launch other consoles such as the MetaFrame
Presentation Server Console.
The Report Center extends Resource Manager reporting capabilities, and allows
you to easily generate reports from a variety of real-time and historical data sources.
A wizard helps you select the type of report, the data to be displayed, and the
schedule for running the report. You can view the status of your scheduled report
and adjust the report parameters before you run it.
Information about installation and software requirements for the Access Suite
Console is in the MetaFrame Presentation Server Administrator’s Guide. User
assistance for the extensions is provided on-screen in the Access Suite Console.
Note that the reports available in Report Center are different from those available in
Resource Manager. Similarly, Dashboard allows you to display performance
metrics differently from Resource Manager. So, to familiarize yourself with all the
monitoring and reporting capabilities at your disposal, explore both the Presentation
Server Console (containing Resource Manager) and the Access Suite Console
(containing Dashboard and Report Center).
The Report Center contains several report types, for example:
• Application Availability report. Lists the published applications in a farm and
the percentage of time when they were available, in a planned down state, and in
an unplanned down state.
• Server Availability report. Lists the servers in a farm and the percentage of
time when they were available, in a planned down state, and in an unplanned
down state.
• Server Performance report. Displays, for the selected servers, the average
CPU utilization, the minimum memory available, and the maximum number of
concurrent connections. The report also contains details of the most heavily
loaded servers.
• Application report. Provides details of all published applications including the
servers they are published on, users and groups with access to them, and any
options used by clients that connect to the applications.
• Application Usage report. Shows the most frequently used applications, the
number of times each application was accessed, and the maximum number of
concurrent users.
• Server Snapshot report. Details the values of performance metrics at a
specified time on any server in your farm.
For further details of how to use the Report Center, see the assistance provided on
the Report Center screens, the Access Suite Console online help system, or the
MetaFrame Presentation Server Administrator’s Guide.
Extended Real-Time Monitoring Capabilities

The Dashboard on the Access Suite Console allows you to display performance
metrics in a highly visual way for the servers in a farm.
You can create custom views of farms, zones, servers, and applications by using the
My Views feature. My Views allows you to create and save customized views of
items in your deployment. You can then see updated information about those items
at any time, without having to search for them. For example, you can group related
objects in folders or you can order farms, according to the frequency with which
you need to access them.
You may want to monitor farms located in different locations. In this case, you can
use the My Views feature to create a custom view that groups each farm in a
different part of the screen and then use a background graphic to identify each
location. A suitable performance metric, such as CPU Load, displayed in your
custom view, allows you to pinpoint problems quickly on any server in any
location.
Using custom views in this way helps you anticipate or identify problems with your
servers or applications as soon as possible. Note that these servers must be running
the MetaFrame Presentation Server Provider for Microsoft Windows Management
Instrumentation.
For further information, see the assistance provided on the Access Suite Console
screens, the Access Suite Console online help system, or the MetaFrame
Presentation Server Administrator’s Guide.
What to Read Next

This guide provides:
• Installation instructions.
• An overview of the tasks that you carry out using Resource Manager. For step-
by-step instructions on how to carry out Resource Manager tasks see the online
help.
• Detailed information on Resource Manager-related tasks that you carry out

using other tools or products.
• Reference information.
To find Refer to
Resource Manager installation intructions Chapter 2: “Installing Resource Manager”
An overview of real-time monitoring Chapter 3: “Monitoring Servers and
Applications in Real Time”
An overview of how to record the history of Chapter 4: “Recording the History of Servers
servers and applications using a summary and Applications”
database
An overview of how to generate reports from Chapter 5: “Reporting and Analyzing Resource
stored Resource Manager information Manager Information”
An overview of how to charge users for Chapter 6: “Billing Users for Resource Usage”
resource usage using summary database
information
Answers to common questions about Resource Chapter 7: “Troubleshooting”
Manager
Details about the default metrics that are Appendix A: “Default Metric Set”
configured during installation
Definitions, layout, and organization of Appendix B: “Summary Database Schema”
summary database schema
Definitions of the technical terms used in this Appendix C: “Glossary”
guide
CHAPTER 2
Installing Resource Manager
Overview
This chapter explains how to install Resource Manager. Topics include:
• Issues to consider before you start, including software requirements and
licensing information
• Installation instructions
• How to upgrade to the latest version of Resource Manager
• An overview of the user interface and navigation
• How to set up a summary database, including setting up a system data source
name (DSN) and database connection server
• How to turn the summary database on and off
• How to change the location of Resource Manager after installation
Software Requirements
You install Resource Manager when you install or upgrade your servers to
MetaFrame Presentation Server 4.0, Enterprise Edition for Windows.
If you are upgrading from a previous release of MetaFrame Presentation Server, it
is not necessary for Resource Manager to have been installed previously.
For guidelines about the licensing requirements for MetaFrame Presentation Server
or other components, see the MetaFrame Access Suite Licensing Guide.
If you are going to record the history of server and application metrics using the
summary database, you need one of the following Database Management Systems
(DBMS) packages:
• Microsoft SQL Server Versions 7 or 2000
—Or—
• Oracle Database Versions 8i, or 9i
Licensing Information
Full Resource Manager functionality requires a MetaFrame Presentation Server
Enterprise Edition license. Refer to the MetaFrame Access Suite Licensing Guide
for further details.
Installing Resource Manager

Before You Start
Before you install Resource Manager, Citrix recommends that you do the
following:
• Ensure that the servers on which you are going to install Resource Manager
meet the software requirements listed in “Software Requirements” on page 17.
Resource Manager stores recent information about applications, servers, and
users in a local database that can become large if the server is heavily loaded. If
this will cause an issue in the future, you can select an alternative installation
location for Resource Manager. See “Installing or Upgrading Resource
Manager” on page 19 for more information.
• Ensure that each server on which Resource Manager is to be installed can
connect to a data store. The data store is a database that MetaFrame Presentation
Server and its components use to keep track of configuration information about
the servers, applications, and configured users in the server farm.
You set up a data store during MetaFrame Presentation Server installation.
Resource Manager uses this data store.
• If you are going to use a summary database, you need to install DBMS software
on a server. This computer is the database storage facility for your server farm’s
summary data. It does not need to be a farm server but must be available to the
farm servers through the network. Ensure that this server has enough available
space to store the summary database. Refer to “Managing the Resource
Manager Summary Database” on page 52.
Go to Document Center Chapter 2 Installing Resource Manager 19
Selecting a Server to Use as the Farm Metric Server

By default, the first server on which you install Resource Manager becomes the
Farm Metric Server. The Farm Metric Server interprets metrics that apply to the
entire server farm (for example, application counts) and sends alerts when required.
Citrix recommends that the Farm Metric Server be lightly loaded and, preferably,
be a data collector. If necessary, you can change the Farm Metric Server to a
different machine after installation. For more information, see the Resource
Manager online help system.
The second server on which you install Resource Manager becomes the backup
Farm Metric Server.
If these servers will experience heavy loading or are not data collectors, specify
different servers to be the Farm Metric Servers.
If You Are Upgrading from a Previous Version

If you decide to upgrade servers to MetaFrame Presentation Server over a period of
time (rather than simultaneously), ensure that you upgrade the Database Connection
Server first, then the main Farm Metric Server, then the backup Farm Metric Server,
before upgrading other servers in the server farm. Problems may occur if another
server is running a later version of Resource Manager.
Installing or Upgrading Resource Manager

Use the following procedure to install or upgrade Resource Manager. You need to
follow this procedure for each server in your server farm.
To install Resource Manager on a server

1. Follow the installation instructions in Chapter 5 of the MetaFrame Presentation
Server Administrator’s Guide, making sure that you choose to install
MetaFrame Presentation Server Enterprise Edition.
2. When the Component Selection Setup screen appears, if you want to change
the installation location of Resource Manager, browse to the correct location
then click OK.
3. At the prompt, click Restart to restart the server.
Note You need to install the Presentation Server Console on every server from
which you want to administer Resource Manager servers. For instructions on how
to install the console, see “To install the Presentation Server Console on a
workstation” on page 27.
Uninstalling Resource Manager

Important If you are uninstalling MetaFrame Presentation Server from the
Resource Manager Farm Metric Server(s) or Database Connection Server for a
summary database, reassign the server before uninstalling. That is, change the Farm
Metric Server(s) and/or the Database Connection Servers to other Resource
Manager servers before uninstalling or removing from the server farm. If you are
using a summary database, Citrix recommends that you update it before removing
any servers from the server farm. For details of how to update the database, see the
Resource Manager online help system.
To uninstall Resource Manager from a server

1. Log off from any currently connected clients and the Presentation Server
Console, and exit all programs running on the server.
2. From the Start menu, select Settings > Control Panel > Add/Remove
Programs.
3. Select Citrix MetaFrame Presentation Server for Windows.
4. Select Change. The Setup wizard starts.
5. On the Application Maintenance screen, select Modify, then click Next.
6. On the Component Selection screen, select Resource Manager and select
Entire feature will be unavailable.
7. Select Next. The selections you have made are listed for review.
8. Click Finish.
Note If you run the uninstaller after manually changing the location of Resource
Manager, as described in “Changing the Location of Resource Manager after
Installation” on page 30, the uninstaller does not remove the Resource Manager
folder, and you must delete it manually.
Setting Up a Summary Database

Before you can start using a summary database, you must do the following:
1. Install your DBMS software on a server and create a database on it.
This server is the database storage facility for your server farm’s summary data.
It does not need to be a farm server but needs to be available to the server farm
through the network. You need to ensure that this server has enough available
space to store the summary database. Refer to “Managing the Resource
Manager Summary Database” on page 52. Resource Manager supports the
following DBMS software:
• Microsoft SQL Server Versions 7 and 2000.
• Oracle Database Versions 8i, and 9i. If you are using an Oracle DBMS,
ensure that the character set it uses contains all the characters you use in
your server farm; for example, for server and application names. This
includes special characters and currency symbols.
Important When you create your summary database on the DBMS server, the
DBMS access credentials you set to be used by Resource Manager must each
not exceed 255 characters in length. This is regardless of the limits of the
DBMS software itself.
If you are using a Microsoft SQL Server DBMS, do not use the master database
for your summary database. The master database is used by SQL Server for
internal functions. Using it for your summary database may cause database
corruption problems.
Citrix recommends that you do not install the DBMS on the Database
Connection Server.
2. Set a system data source name (DSN).

The system DSN stores information about how a client can connect to a
database. It is required by the Database Connection Server (the database client)
to be able to communicate with the summary database DBMS. Refer to “Setting
a System Data Source Name” on page 22.
3. Configure a Database Connection Server.
This server enables communications between the server farm and the summary
database by writing data to the database and reading data from it. It should be
relatively low-load for best performance. Refer to “Configuring a Database
Connection Server” on page 24.
Setting a System Data Source Name

To set a system Data Source Name for Microsoft SQL Server DBMS
1. Choose a server to be your Database Connection Server.
2. Open the Windows Control Panel.
3. Open the ODBC Data Source Administrator dialog box. To do this, open
Administrative Tools, then open Data Sources (ODBC).
4. Click the System DSN tab.
5. Click Add.
6. In the Create New Data Source dialog box, select SQL Server.
7. Click Finish.
8. In the Microsoft SQL Server DSN Configuration dialog box, type
rmsummarydatabase in the Name box. Type a description (optional), then
select the server with the DBMS installed on it from the Server list.
Important You must type rmsummarydatabase exactly. Any spaces or

spelling errors will make the database unrecognizable to the Database
Connection Server. However, the field is not case-sensitive.
Screenshot of the Microsoft SQL Server DSN Configuration dialog box

9. Click Next.
10. Select how Microsoft SQL Server authenticates your identification so you can
set up the system DSN. Either:
• Click With Windows NT authentication using the network login ID to
use Windows NT authentication
—Or—
• Click With SQL Server authentication using a login ID and password
entered by the user, then select the Connect to SQL Server to obtain
default settings for the additional configuration options check box and
type a user name and password in the Login ID and Password boxes,
respectively
11. Click Client Configuration.
12. In the Edit Network Library Configuration dialog box, select TCP/IP under
Network libraries.
13. Click OK.
14. Click Next.
15. Select the Change the default database to check box, then select the database
you created on the DBMS server from the list.
16. Click Next, then Finish.
17. In the ODBC Setup dialog box, you can click Test Data Source to confirm the
DSN configuration. Click OK twice to close the dialog box.
18. Click OK to close the ODBC Data Source Administrator dialog box.
For more information, refer to your Windows operating system and Microsoft SQL
Server documentation.
To set a system Data Source Name for Oracle DBMS
Note The system DSN setup described below may be different for other Oracle
versions.
1. Choose a server to be your Database Connection Server.

3. Open the ODBC Data Source Administrator dialog box. To do this, open
Administrative Tools, then open Data Sources (ODBC).
4. Click the System DSN tab.
5. Click Add.
6. In the Create New Data Source dialog box, select the Oracle ODBC Driver
option. This option is available only after the Oracle Client is installed.
7. Click Finish.
8. In the Oracle ODBC Driver Configuration dialog box, type
rmsummarydatabase in the Data Source Name text box. Type a description
(optional).
Important You must enter rmsummarydatabase exactly. Any spaces or

spelling errors will make the database unrecognizable to the Database
Connection Server. However, the field is not case-sensitive.
9. From the TNS Service Name list, select the global database name of the Oracle
database and type the user name in the User ID box.
10. Select Disable Microsoft Transaction Server (MTS) on the Workarounds
tab. For information about why MTS needs to be disabled, refer to Microsoft
Knowledge Base articles Q180190 and Q193893, available from
http://www.microsoft.com.
11. Ensure that:
• The Read Only check box is cleared
• Enable closing cursor and Enable Results Set are selected
12. Click OK to close the Oracle ODBC Driver Configuration dialog box. Note
that the dialog box shown varies between Oracle releases.
For more information, refer to your Windows operating system and Oracle
Database documentation.
Configuring a Database Connection Server

After you set up a system DSN on the Database Connection Server, you need to
configure this server as the Database Connection Server using the Presentation
Server Console.
To configure a Database Connection Server

1. In the left pane of the console, click Resource Manager.
2. In the right pane, click the Summary Database tab.
3. Click Configure.
4. In the Summary Database Configuration dialog box, select your Database
Connection Server from the Server list. Only servers running Resource
Manager appear in the list.
Note If a server name is dimmed, this means that the server is running an older
version of Resource Manager and should not be selected.
5. Enter the DBMS access credentials in the User and Password boxes. These
must match valid credentials defined within the supporting DBMS (the Oracle
or Microsoft SQL Server database you are using).
Note Resource Manager supports Windows NT authentication for the

Microsoft SQL Server user name. The 255 character limit for the user name
includes the domain name, the intervening backslash ( \ ), and the user name.
6. Click Test to check the connection to the database.

You can now activate the summary database. See “Turning the Summary Database
On” on page 25.
You can further configure the Database Connection Server in the following ways:
• Configure the database update time. See “Setting the Summary Database
Update Time” on page 48.
• Configure a database purging schedule to remove unwanted information. See
“Removing Unwanted Information from the Database” on page 47.
• Choose the methods for sending alerts when an update to the database fails. See
“Preparing Your System for Resource Manager Alerts” on page 37.
• Configure data collection restrictions for all farm servers. See “Using Farm-
Wide Options for Reducing Summary Data” on page 46.
Turning the Summary Database On

To begin recording data for your database, you need to turn the summary database
on after installation. When it is on, Resource Manager servers create and store
information for inclusion in the summary database.
To turn the summary database on

1. In the left pane of the Console, click Resource Manager.
Note The first icon in the Status panel is Not Configured when the
summary database is off or a Database Connection Server is not configured. In
this state, Resource Manager servers are not creating or storing information for
inclusion in the summary database.
3. Click Configure.
4. In the Summary Database Configuration dialog box, select the Summary
Database enabled check box.
5. Click OK twice.
The first icon in the Status panel is OK , meaning the summary database is
on and a Database Connection Server is correctly configured and in use. In this
state, Resource Manager servers are collecting information for inclusion in the
database.
Turning the Summary Database Off

If you need to stop creating summary data, for example for maintenance purposes,
you can turn the summary database off.
CAUTION When the summary database is off, Resource Manager servers will no
longer summarize and store information for the summary database. This may result
in data loss until the summary database is turned back on.
You cannot turn the summary database off for individual servers; however, you can
minimize the data being contributed by choosing to ignore server metrics for
periods of low activity for an individual server, for example, weekends or late at
night. Refer to “Ignoring Server Metrics during Periods of Low Server Activity” on
page 47 for details.
To turn the summary database off

1. In the left pane of the Presentation Server Console, click Resource Manager.
3. Click Configure.
4. In the Summary Database Configuration dialog box, clear the Summary
Database enabled check box.
5. Click OK twice.
The first status icon in the Status panel is Not Configured , meaning the
summary database is off. In this state, Resource Manager servers are not
summarizing or storing information for inclusion in the summary database.
Displaying Resource Manager and its Components

The following procedures are designed to familiarize you with the interface so that
you can quickly get up to speed with Resource Manager. The user interface for
Resource Manager is integrated with the Presentation Server Console.
Note Extended reporting and monitoring functionality available from the Access
Suite Console (Report Center). For more information, see the assistance provided
on the Access Suite Console screens, the Access Suite Console online help, or the
MetaFrame Presentation Server Administrator’s Guide.
To open the Presentation Server Console

From the Start menu, choose Programs > Citrix > MetaFrame Presentation
Server > Presentation Server Console. You can also open the Presentation Server
Console from the Access Suite Console if the Access Suite Console is appropriately
configured.
When the console starts, log on to a server.
When you are connected to a server farm, the console displays a window with two
main panes:
• The left pane shows a hierarchical list of the components of the server farm
• The right pane shows information about the object that is selected in the left
pane
From this window you can access Resource Manager. The servers and applications
that you see depend on whether you are a full administrator or a custom
administrator. If you are a full administrator ,you have access to the entire server
farm. If you are a custom administrator, you can view and update only those folders
of servers or applications for which the full administrator has granted you
permissions.
You need to install the console on every machine from which you want to
administer servers with Resource Manager installed. You can install the console on
a MetaFrame Presentation Server computer (at the time when you install
MetaFrame Presentation Server itself), or on a remote machine.
To install the Presentation Server Console on a workstation

1. Ensure that Resource Manager is installed on the server for which you want to
view Resource Manager information.
2. Install or upgrade the Presentation Server Console on the workstation using the
Setup program on the MetaFrame Presentation Server CD-ROM.
Displaying the Main Resource Manager Screen

To display the main screen
In the Presentation Server Console, either:
• In the left pane, click Resource Manager.
—Or—
• In the right pane, double-click Resource Manager.
Screenshot of the main Resource Manager screen

This window displays a number of tabs that enable you to perform the following
functions in Resource Manager:
• Watcher. Show a real-time list of all servers in the server farm that have an
alarm state.
• Reports. Generate reports about:
• Current process and user activity, and recent server status.
• Past process, user, and server activity. These reports require a summary
database to be in use.
• Summary Database. Configure a summary database and see its status.
• Billing. Configure cost centers (fees and user groups). If you are using a
summary database, generate Billing reports based on resource usage.
• SMS, SNMP, and E-mail. Configure automatic SMS, SNMP, or email alerts.
• Farm Metric Server. See the status of Farm Metric Servers and change the
servers being used as Farm Metric Servers.
Displaying Resource Manager for the Entire Server Farm

This view enables you to monitor all the servers in your server farm. You can gain
an overall picture of the status of the server farm and spot problems as they occur.
To display Resource Manager for the entire server farm

1. In the left pane of the console, click Servers.
2. In the right pane, click the Resource Manager tab.
Displaying Resource Manager for a Single Server

To display Resource Manager for a single server
1. In the left pane of the console, navigate to the required server.
2. In the right pane, click the Resource Manager tab.
This displays all the metrics that are being monitored for that server. If a
problem arises, a status icon appears to warn you.
Displaying the Applications in a Server Farm

Monitoring in a server farm
In the left pane of the console, expand the Applications folder.
You can use Resource Manager to monitor all published applications that are
running on Resource Manager servers in the server farm. When you start using
Resource Manager, all existing published applications are listed in the Applications
folder.
You can also monitor applications that are not published by setting them up as
Resource Manager applications, and identifying the servers on which you want to
monitor them. For details of how to set up Resource Manager applications, see the
online help system. When you set up a Resource Manager application, it is added to
the list in the Applications folder. Refer to “Monitoring Your Farm” on page 36.
Changing the Location of Resource Manager after Installation

If necessary, you can change the location of Resource Manager after installation.
Before changing the location of the Resource Manager files, you must ensure that
you set the correct permissions so that the Independent Management Architecture
(IMA) service can read and write files at the new location.
CAUTION This procedure requires you to edit the registry. Using Registry Editor
incorrectly can cause serious problems that may require you to reinstall your
operating system. Citrix cannot guarantee that problems resulting from the incorrect
use of Registry Editor can be solved. Use Registry Editor at your own risk. Be sure
you back up the registry before you edit it.
To change the location of Resource Manager

1. Stop the IMA service.
2. Move the Resource Manager folder and all its contents to the new location.
3. Edit the registry key InstallDir in
HKEY_LOCAL_MACHINE\SOFTWARE\Citrix\Citrix Resource Manager
4. Restart the IMA service.
Note If you run the uninstaller after manually changing the location of Resource
Manager, as described here, the uninstaller does not remove the Resource Manager
folder, and you must delete it manually.
CHAPTER 3
Monitoring Servers and

Applications in Real Time
Overview
Resource Manager provides you with information about a number of system and
network processes and events. Status displays show this information in real time,
enabling you to see the state of your system at a glance. You can monitor the
following:
• The status of the servers in a server farm
• The number of instances of specific applications that are running in a server
farm
Each item that is being monitored is referred to as a metric.
A metric is a combination of:
• The type of object that you want to monitor: a physical or logical system
resource; for example, a computer's hard drive.
• The counter to be monitored: the specific aspect of the object that you want to
monitor; for example, disk free space.
• The instance of the object: an individual example of the object or a state it needs
to reach to be counted. For example, a computer may have more than one hard
drive. In this case, the instance identifies which drive you want to examine.
When you install Resource Manager, a default set of metrics is configured
automatically for each server. You can change the metrics to suit your specific
environment. A set of default limits is also configured during installation for the
metrics that apply to each server. You can change these limits to suit your needs.
When a metric’s value exceeds its defined limits, Resource Manager displays a
warning or problem status icon for the metric. These are known as alarms.You can
also configure Resource Manager to send messages to notify you of warnings and
problems. These are known as alerts.
Note In this release of Resource Manager, the speed of alert notification is much
faster.
Each server with Resource Manager installed has a Microsoft Jet Access database
in which it stores metric values and application information for the last 96 hours. By
default, this database is located in:
...\Citrix Resource Manager\LocalDB\RMLocalDatabase. It is accessed when you
are creating real-time graphs, displaying Server Snapshot reports, and running
reports on that specific server. The IMA service reads and writes to the database
periodically. The size of the database is managed automatically. When the IMA
service is started on a server, the local Resource Manager database is compacted
every day automatically.
You can also use the Management Console for MetaFrame Access Suite to display
performance metrics for the servers in a farm in a highly visual way, as described in
“Extended Real-Time Monitoring Capabilities” on page 15. For further
information, see the assistance provided on the Access Suite Console screens.
Configuring Metrics on Servers

When you install Resource Manager, it automatically configures a default set of
metrics for each server. These default metrics are described in Appendix A.
Resource Manager metrics are derived from Microsoft Windows Performance
Monitor; Resource Manager can track any Windows Performance Monitor counter
as a server metric. For explanations of the metrics and advice on customizing
metrics, refer to Appendix C of the MetaFrame Presentation Server
Administrator’s Guide and the Microsoft Windows Help for Performance Monitor.
Resource Manager configures default alarm thresholds for the default metrics.
Citrix recommends that you customize these metrics over time to suit your
environment, and that you limit the total number of metrics being tracked on a
server to 50.
The default metrics provide a real-time overview of each server but, to performance
tune each server, you can add more specific metrics. This set of metrics is, by
default, recorded in the summary database if one is configured.
For further information on how to configure metrics, see the Resource Manager
online help system.
Go to Document Center Chapter 3 Monitoring Servers and Applications in Real Time 33
Checking the Status of Server Metrics

For each server in your system, status icons show the status of each monitored
condition or metric.
During installation, Resource Manager automatically configures a set of limits for
the metrics that apply to each server. If a metric’s value falls outside normal limits
for a particular length of time (a tolerance time that you can configure), the status
icons change. You may need to alter these limits to suit your specific environment.
You can set alerts to be triggered when either yellow (warning) or red (problem)
status occurs, and also when an alert state returns to green (OK). See the online help
system for more information on setting metric limits.
You can also set alerts to be sent when the status of a metric returns to green (OK)
or yellow, having been previously at yellow or red. The meaning of each status icon
is described below:
Represents normal operation; that is, the value of the metric falls between the
OK set limits.
Represents a warning condition. This means that a problem may be

Warning developing that will require further analysis to improve performance or to
prevent the situation from becoming worse.
Represents a problem condition. This often means that some action is required
Critical to provide better application or server performance.
Both yellow and red indicators occur when the value for a metric falls outside
the normal limits and remains there for a defined period of time.
Represents a metric that is not yet active and needs to be configured.
Not
Configured
Represents a metric that is set to “Sleep;” that is, you indefinitely suspended
Sleep notification of the metric’s status. See “Suspending Notification of a Metric’s
Status” on page 37 for details. This icon is also used (for all metrics) if the
server is unlicensed.
Represents a metric that is set to “Snooze;” that is, you suspended notification
Snooze of the metric’s status for a fixed period. See “Suspending Notification of a
Metric’s Status” on page 37 for details.
Resource Manager determines the status of each metric by sampling the raw
performance data every 15 seconds and updates metric status icons accordingly.
The history of metrics being monitored on a server is recorded in the Resource
Manager server log. For details of how to view the log, see the help system.
Customizing Server Metrics

If you need to change the way in which Resource Manager monitors the servers in
real time, you can:
• Change the list of metrics being monitored for a server
• Configure the way alerts work for individual metrics
• Change the list of processes that Resource Manager does not monitor
If you need to configure a new metric, you can display a graph of the current values
for that metric using the Visual Threshold Configuration option, and use this to help
you set the appropriate thresholds.
You can also change the server that deals with farm-wide metrics. In the server
farm, the Farm Metric Server interprets metrics that apply to the entire server farm
(such as application counts) and sends alerts if necessary.
Running a Script on Transition to Red

You can configure an executable file to run on a local Resource Manager server
once a server metric indicates that a problem condition exists. The script you select
runs under the System account, and launches every time the metric enters a red
alarm state (not just the first time). It must, therefore, be one that can run as an
automated response to alarms (that is, without a user interface) and it must
terminate automatically.
To configure a script to run on transition up to red

1. In the left pane of the Console, expand the Servers folder, then click the name
of the local server for which you want to add a script.
2. In the right pane, click the Resource Manager tab to display the current status
of the server.
3. Select the metric for which you want to run the script, right-click and select
Properties. The Server Metric Properties dialog box opens.
4. Click the Script configuration tab.
5. Click Run Script on transition up to red.
6. Click Browse to locate the script. The script must be on a local drive and the
System account must have access to run it. Select only executable files here and
be aware that they run as a process in the background.
Note You cannot type directly in the box, you must click Browse.
7. Click Apply to other servers to copy metric configuration to other servers or

applications.
8. To copy the configuration of the current metric to other metrics in the Metrics
list, click Copy properties to other metrics.
9. Click OK to confirm your changes and Resource Manager runs the executable
file when the metric indicator you specified turns red.
Monitoring the Disk on a Windows 2000 or 2003 Server

By default, Resource Manager monitors the LogicalDisk performance object on
Windows 2000 and 2003 Server machines.
Some installations of Windows 2000 Server, without Windows 2000 Service Pack 1
installed, do not include the LogicalDisk performance object. If you have this type
of installation, no disk monitoring metric is included in the default list of metrics.
You can address this in two ways:
• You can monitor the PhysicalDisk performance object instead by adding
appropriate PhysicalDisk metrics using Resource Manager. This means that
only the physical drive is monitored, not each partition, and % Disk Free Space
is not monitored.
• You can install Windows 2000 Service Pack 1, which enables you to monitor
the LogicalDisk object. This is preferable, because you can then also monitor
the physical drive, partition, and % Disk Free Space.
Configuring Application Metrics

Resource Manager monitors only one metric for each application. This is the Count
metric, which maintains a count of how many instances of specific applications are
running in the server farm, and notifies you if the number of instances of a
monitored application reaches a defined limit. This enables you to manage
application licenses.
You can monitor all published applications. You can also monitor applications that
are not explicitly published, for example, because the desktop itself is published, or
because you are not deploying the application through MetaFrame Presentation
Server. To do this, you need to set up the application as a Resource Manager
application. See the Resource Manager online help system for more information.
Resource Manager can monitor a published application only if you specified the
full path name of the application in the Properties dialog box when you published
the application in MetaFrame Presentation Server. When you publish an
application, it is a good idea to use the Browse button to select the executable to
ensure that you use the correct full path. For full information on publishing
applications, refer to the MetaFrame Presentation Server Administrator’s Guide.
If an application path name also specifies command-line parameters for the
program, surround the path name with double quotes (“ ”). This enables Resource
Manager to distinguish between process name and argument list and monitor the
published application correctly.
Important You can use Resource Manager only to keep track of the usage of 32-bit
applications. You cannot monitor 16-bit applications.
Monitoring Your Farm

Resource Manager enables you to monitor the status of servers and applications in
your server farm in a number of ways.
You can organize servers and applications in the server farm in folders and use
Resource Manager to monitor each folder as a unit. You can then see an overview
of the servers or applications in the folder, as well as an overview of the entire
server farm. See the online help system for more information on real-time
monitoring.
If you do not want to show the full status display on the screen, you can monitor
servers or server folders by displaying a smaller window, called the Watcher
Window.
Resource Manager can send alert messages to notify you under the following
conditions:
• A server in the server farm unexpectedly stops operating
• The status of selected metrics changes:
• Up to red
• Up to yellow
• Down to yellow
• Down to green
Alerts are sent by email, SNMP message, SMS message, or more than one of these
formats. The Farm Metric Server interprets metrics that apply to the entire server
farm (for example, application counts) and sends alerts when required.
Alerts are generated whenever the IMA service stops operating, including expected
events such as scheduled reboots.
Suspending Notification of a Metric’s Status

You can stop Resource Manager from giving you information about a specific
metric. This is useful when, for example, you want to work on a problem without
receiving repeated alarms. When you suspend notification, Resource Manager
continues recording information about the metric's values but does not display
alarms or send alerts. See the online help system for more information.
Preparing Your System for Resource Manager Alerts

You can use Resource Manager to send alert messages when metrics change state.
The alerts can be sent using one or more of the following methods:
• Email messages, using MAPI or SMTP
• Short Message Service (SMS) text messages to cell/mobile phones
• Simple Network Management Protocol (SNMP) messages
To use Resource Manager to send alerts, set up one or more Resource Manager
servers to send the alert messages. Make sure that these servers have additional
hardware or software to handle each type of alert you require; for example, a
modem for SMS alerts and an email system for MAPI alerts. For SNMP alerts, you
need to set up SNMP on every server that has one or more metrics configured to
send SNMP alerts.
Note After you install the SNMP service, restart the machine for Resource
Manager alerts to be sent or received.
.
When you have done this, you can configure settings in Resource Manager, such as
who will receive the alert messages, and set up the individual server and application
metrics that you want to trigger the alert messages.
The alert recipients you set up will be used for all the servers in the server farm. You
can modify Resource Manager alert recipients for any individual server.
If you want delegated responsibility for a set of servers to a custom administrator,
you may wish to prevent that administrator from receiving alerts from all the other
servers in the farm. You do this by setting up the administrator’s alert contact details
in the Add MetaFrame Presentation Server Administrator wizard. For each server
folder, you can specify whether or not the custom administrator will receive alerts
for the servers in that folder. For further details, see the online help system provided
with the wizard.
Preparing Your System for Email Alerts

Before you can use email alerts, you must configure your system.
You can use either MAPI or SMTP to generate email alerts. Select the relevant
setting in the Resource Manager Properties dialog box. By default MAPI is used to
send alerts.
• For MAPI email alerts, you first need to choose and configure the servers in the
server farm that you want to use to send the email messages. Then you create a
mail profile and enable the Resource Manager MAPI email service. These
servers are called MAPI Connection Servers.
• For SMTP email alerts, you must have an SMTP email server that is accessible
from your server farm. You then specify the details of your SMTP server in
Resource Manager. See the Resource Manager online help system for more
information.
You can secure communication between your Resource Manager servers and
your SMTP server using SSL. For details of how to set up SSL, see “Setting up
SSL for SMTP Email Alerts” on page 40.
Using MAPI to Send Alerts

To use MAPI to send alerts, you must:
• select the MAPI Connection Servers
• create a mail profile for Resource Manager
• enable the Resource Manager MAPI Mail service
• configure Resource Manager to use MAPI email alerts
Selecting the MAPI Connection Servers

You need to select one or more servers in the server farm to be the MAPI
Connection Servers. When email alerts are generated in the farm, they are passed to
one of the configured MAPI servers, which then sends the actual email. You can
configure as many MAPI Connection Servers as you wish. Alerts are sent through a
randomly selected MAPI server, although if the server on which an alert is
generated is also a MAPI server, that server sends the email.
Each MAPI Connection Server must be able to access a mail server (for example,
Microsoft Exchange Server). It must have an email client installed (for example,
Microsoft Outlook) that conforms to the X-400 protocols.
Creating a Mail Profile for Resource Manager

On each MAPI Connection Server, you need to configure a mail profile for
Resource Manager to use. The profile must have the same name and details on all
of your MAPI Connection Servers. Citrix recommends that you give the profile a
name that is easy to recognize (for example, Resource Manager). The profile is
used throughout the server farm.
When you create the mail profile, ensure that you include the mail system that you
want to use; for example, Microsoft Exchange Server. You can also specify an
address book for the profile.
After you have set up the profile, it is a good idea to test that you can log on to your
email system using the profile and that you can send a message.
Note A user who is configured to use the email profile can log on to the email
system without being prompted for logon credentials.
For more information on configuring email profiles, refer to Citrix Knowledge
Base article CTX333658 available at http://knowledgebase.citrix.com.
Enabling the Resource Manager MAPI Mail Service

Email alerts are managed by a service called Resource Manager Mail. This service
is installed automatically on all servers on which you install Resource Manager.
You need to enable the service on each MAPI Connection Server.
To enable the Resource Manager Mail Service on a MAPI Connection Server

2. Open the Services dialog box. To do this, open Administrative Tools, then
open Services.
3. In the Properties dialog box for the Resource Manager Mail service, click the
General tab and ensure that the startup type for the service is set to
Automatic.
4. From the Log On tab, select This account.
5. Enter the details of the local user account, including the domain, that you want
Resource Manager to use for email alerts. Ensure that you type the account and
domain details exactly, or browse to the account so that you can be sure that the
details you enter are correct.
6. Ensure that the Resource Manager Mail service is started.
Configuring Resource Manager to Use MAPI Email Alerts

After you set up the MAPI Connection Servers, you must set up Resource Manager
to use email alerts for the server farm. For example, you need to specify the alert
recipients. These settings apply to the entire server farm. See the Resource Manager
online help system for more information on MAPI email alerts.
Setting up SSL for SMTP Email Alerts

Important This section assumes that you have access to a certificate authority and
an appropriate root certificate, and are familiar with the procedure of adding a
certification path to a Windows server. You must also ensure that your SMTP server
supports SSL and is properly configured to use SSL in your network environment.
1. Install the certification path of your certificate authority on each Resource

Manager server. In the Microsoft Management Console, add the root certificate
path in the following location:
Certificates (Local Computer)\Trusted Root Certification
Authorities\Certificates
This step enables each server to trust certificates issued by your certificate
authority.
3. In the right pane, click E-mail. From the E-mail tab:
1. Specify the SMTP server by typing its fully qualified domain name
(FQDN). For example:
smtpserver1.mydepartment.mycompany.com
2. Click Use SSL to send e-mail alerts.
3. Click OK.
Preparing Your System for SMS Alerts

To send an SMS message to a cell/mobile phone, a Resource Manager server makes
a call through its modem to a number that is designated by the cell/mobile service
provider. It then sends data to the service provider's computer, instructing it to send
an SMS message to the cell/mobile phone of the person who is to receive the alert.
If you want to use SMS alerts, ensure that at least one Resource Manager server in
the server farm has a modem. This can be an analog modem or an ISDN card.
You need to investigate the modem requirements of the service providers for the
cell/mobile phones to which you want to send alerts. Some service providers
require a specific type of modem (usually analog). Where this is the case, at least
one server with that type of modem must be in the server farm before you can use
SMS alerts for that service provider.
If the people that you want to receive SMS alerts use a variety of service providers,
you need to know the details of the gateway that Resource Manager must use to
communicate with each service provider. Each provider is likely to have a different
telephone number, and may employ a different protocol to carry the messages.
Some service providers offer an analog line, others offer ISDN. You probably need
to configure a range of numbers to call and a range of protocols to use.
The alert recipients you set up will be used for all the servers in the server farm. You
can modify Resource Manager alert recipients for any individual server.
Configuring Resource Manager to Use SMS

When you are sure that you have fulfilled all the requirements for SMS alerts, you
can configure the way in which Resource Manager uses SMS for alerts in the server
farm. For example, you need to specify the alert recipients. These settings apply to
the entire server farm. You can also specify SMS alert recipients for individual
servers.You need to select one or more servers in the server farm to send the SMS
alerts. Such a server is called a TAPI Server. For more information on configuring
Resource Manager to use SMS, see the Resource Manager online help system.
Preparing Your System for SNMP Alerts

Resource Manager can send five different SNMP alerts:
• trapServerDown - The Resource Manager server is down
• trapMetrictoGreen - The metric on the Resource Manager server has changed
to green status
• trapMetricGreenToYellow - The metric on the Resource Manager server has
changed from green to yellow status
• trapMetricRedtoYellow - The metric on the Resource Manager server has
changed from red to yellow status
• trapMetricToRed - The metric on the Resource Manager server has changed to
red status
To receive SNMP messages, a computer on a network requires an SNMP
management tool that enables it to listen for messages. A number of third-party
commercial tools are available.
You need to install and set up the Windows 2000 or Windows Server 2003 SNMP
service on every server that has a metric configured to send SNMP alerts.
Note After you install the SNMP service, you must restart the machine for
Resource Manager alerts to be sent or received.
If you want Resource Manager to send alerts for the Count metric on any
.
applications that are running in the server farm, ensure that the Farm Metric Server
has the Windows SNMP service enabled and running.
Getting More Information About Metrics and Monitoring

Resource Manager uses server performance and resource metrics derived from the
Microsoft Windows Performance Monitor. For explanations of the metrics and
advice on customizing metrics, refer to Appendix C of the MetaFrame Presentation
Server Administrator’s Guide and the Microsoft Windows Help for Performance
Monitor.
CHAPTER 4
Recording the History of Servers

and Applications
Overview
This chapter explains how you can use Resource Manager to store details of server
performance, application instances, and resource usage in a summary database.
Topics include:
• How Resource Manager information gets into a summary database
• What information you should record in a summary database
• Scheduling data collection for a summary database
For information on how to set up a summary database, see “Setting Up a Summary
Database” on page 21.
Resource Manager and the Summary Database

Each Resource Manager server creates a summarized version of its daily activity.
This information is known as summary data. There are various types of summary
data:
• Server-specific performance metrics
• Server-specific session information
• Farm-wide application metrics
• Farm-wide server events (for example, server-down)
Farm-wide metric and server event information is generated as summary data by
the Farm Metric Server in addition to its own server-specific information and
metrics. Farm-wide metrics are routed from servers to the Farm Metric Server
through the zone data collector.
Summary data is kept in special temporary summary files that are stored locally on
each server in a database-compatible format. Each hour, Resource Manager adds
the summary data gathered over the previous hour to the summary files. Summary
files are stored in the following folder: \\Program Files\Citrix\Citrix Resource
Manager\SummaryFiles folder.
On a daily basis, the summary data held by each server in the server farm is
collected by the Database Connection Server. The Database Connection Server then
updates the summary database. After the summary database is updated, summary
files are overwritten with new data.
The following diagram represents a typical server farm utilizing the Resource
Manager summary database. It shows the flow of both server-specific and farm-
wide summary data from the farm servers to the summary database through the
Database Connection Server.
Server Farm
Farm Metric
Server
An Application
An application example
for the purposes
of demonstrating the
having a Summary
Database.
Database
Connection
Server External
An Application
for the purposes
having a Summary
Database.
Database
Server
An Application
for the purposes
having a Summary
Database.
Server
Server
Legend
An Application
Farm-wide metrics
for the purposes
having a Summary
Database.
Server-specific metrics
Farm-wide metrics collated on the Farm Metric
Server continuously (via zone data collector)
Hourly summary data sent to Database
Connection Server once per day
Daily summary data
A diagram showing a server farm using a summary database.

Go to Document Center Chapter 4 Recording the History of Servers and Applications 45
What Information Should I Record in the Database?

Resource Manager servers, by default, automatically record process-related and
session-related information and server events in the summary database. The scope
and detail of this information cannot be changed. A number of default metrics are
also included in the database when you install Resource Manager. You can modify
the default metric set or specify your own for individual or multiple servers.
Note When you make changes to the summary database configuration, it can take
up to 10 minutes for other Presentation Server Consoles in the farm to be updated
with the new settings.
Issues to Consider when Selecting Database Information

You need to consider some important issues when selecting information to be stored
in your database. These are outlined below.
The Benefits of Keeping Long-Term Resource Manager

Information
Store information that will be useful; for example, if you are billing users for RAM
usage, you can store process information until you create bills for it.
It is a waste of summary database space to store information that is of little
significance for your server farm. For example, there is little point recording the
amount of application usage on a server if you don’t care what specific applications
are being used.
Deciding which Metrics Information to Store

You need to decide what metric information to store for different servers in your
server farm. This is important if you have several servers contributing differing
metric information to the database because it may confuse an assessment of the
server farm as a whole when using summary database reports.
Consider the following scenario:
An administrator has selected the “LogicalDisk - % Free Space” metric to be
recorded in the database for half of the servers in the server farm.
At a later date, an administrator generates a report from the summary database to
find the servers most in need of hard drive upgrades (least hard drive free space).
Because half of the servers in the server farm are not storing this metric in the
database, the administrator does not get an accurate report of the overall state of the
server farm for the metric.
How Can Recording a Metric Affect Database Growth?

You need to consider how quickly your summary database will grow when storing
your chosen metric information, and the amount of hard drive space available to the
DBMS. The larger your server farm and the more information you store, the faster
your database will grow. For more information about estimating the growth of the
database, see “Estimating Summary Database Growth” on page 53.
To help you manage database growth, you can remove unwanted data using a
purging schedule. Database purging automatically deletes records older than a
specified age. See “Removing Unwanted Information from the Database” on page
47 for details.
Selecting Server and Application Metrics to Record in the

Database
Whenever you add a new metric, it is automatically set to be summarized. Refer to
Chapter 3 for details about how to set and use metrics.
You can select server or application metrics for individual servers or applications,
or use one setting for multiple servers or applications. See the Resource Manager
online help system for more information.
Using Farm-Wide Options for Reducing Summary Data

Resource Manager enables you to reduce the amount of farm-wide summary data
being generated. You can use the Collection Restrictions option described in the
Resource Manager online help system to create summary data only for processes
connected to, or part of, published applications or Resource Manager applications.
Note Collection restrictions may affect resource billing because process data that
is not stored in the summary database cannot be billed.
Scheduling Summary Data Collection and Removal

Resource Manager gives you control over several important aspects of maintaining
historical records in your summary database. These include:
• Ignoring periods of low server activity
• Automatically deleting records from the summary database by scheduling their
removal after they are stored for a specified time
• Setting when you want the Database Connection Server to update the summary
database with the day’s summary data
The detailed procedures to follow if you want to carry out these tasks are described
in the Resource Manager online help system. The following sections provide a
general overview of the tasks.
Ignoring Server Metrics during Periods of Low Server

Activity
You can configure Resource Manager to ignore server metrics during periods of
low server activity; for example, over weekends or late at night, where metric
collection may be of little benefit and would increase database size unnecessarily.
You can exclude server metrics from summary data on specific days of the week
and/or during specific periods during each day. Alternatively, if you need to capture
data about business processes such as overnight backups, you may prefer to set up
continuous data capture. Note that session, process, application, and server event
information is always recorded regardless of any settings you make.
You can schedule server metric summary data collection for individual servers or
use one setting for multiple servers. Any previous settings for individual servers are
replaced with the new settings.
Removing Unwanted Information from the Database

Over time, your database will grow in size as it stores summary data for the server
farm. Records you store in the database are kept indefinitely, by default, which can
lead to a large database in a short period of time. The greater the number of servers
contributing to the database, and the greater the amount of summary data being
stored for each server, the faster the database will grow.
You may want to keep some of the records stored in the database only for a certain
length of time. For example:
• If you record the percentage of CPU interrupt time so you can assess potential
hardware problems or server overloading on a monthly basis, you may not
require the information after assessment
• If you record session and process information on your servers so you can bill
users for their usage time, you may not want to keep the records after the bill is
created
You can remove unwanted data from the database using a purging schedule.
Purging automatically deletes records from the database after they are there for a
specified period of time. You can also configure data so that it can be purged only
after a bill is created for it. You can configure database purging using the following
record type options:
• Events
• Metrics
• Sessions/processes (billed)
• Sessions/processes (not billed)
Setting a Purging Schedule

When you configure your purging schedule, you must specify how long the data is
to be kept before being purged. The age of a piece of data is calculated by
subtracting its end time from the current Database Connection Server time.
Retention periods are set to “Indefinite” upon initial setup. The Indefinite setting
never purges the associated data.
Important Ensure that the operating system time and date on the originating
Resource Manager servers and the Database Connection Server are synchronized.
This prevents data from being purged incorrectly; for example, if one of the server
farm servers has the date set two days behind that of the Database Connection
Server, the data from it is purged two days earlier than expected.
Setting the Summary Database Update Time

The Database Connection Server automatically updates the summary database with
data from each Resource Manager server once per day. This is referred to as the
update time. The default update time on setup is 00:00 hours (midnight). If this is
inconvenient, change this to a time of day when server activity is low to prevent
slow data transferal or interference with normal server farm activities.
During an update, each server in the server farm first sends a request to the
Database Connection Server asking for permission to send its summary data for the
day. After accepting the request the Database Connection Server receives the
server’s summary data, and then updates the summary database.
Note The update time is always interpreted in the time zone local to each server.
Servers in different time zones will request to send their summary data at the update
time in their local time zone.
You can perform manual updates independently of the update time. You may want
to do this, for example, if you want to generate reports on a fresh set of information.
If an update is unsuccessful, this is reported on the Summary Database tab and an
alert is sent.
You can also temporarily “Sleep” the Database Connection Server to stop the
database from being updated. You may want to do this to perform maintenance on
the database.
Monitoring the Status of the Summary Database

You can easily check the status of the summary database using the Summary
Database tab.
To view the summary database status and run-time activity icons

2. In the right pane, click the Summary Database tab. The Status area includes
indicators showing the current Database Connection Server, Farm Metric
Servers, run-time process, and database import states.
Screenshot of the Summary Database status panel

The meanings of the status icons for the Database Connection Server
On/Off/Configuration indicator are outlined below.
The summary database is off or a Database Connection Server is not

Not configured. In this state, no Resource Manager servers are creating or storing
Configured summary data for inclusion in the database.
The summary database is on and the Database Connection Server is correctly

OK configured. In this state, Resource Manager servers are creating and storing
summary data for inclusion in the database.
The summary database software version for the Database Connection Server
Critical is not accepted because another server in the server farm has a later version
of the summary database software installed. You need to upgrade the
software on the Database Connection Server. See Chapter 2, “Installing
Resource Manager” for details.
Note: This situation can arise only when there are mixed releases of the
summary database software within a server farm.
The meanings of the status icons for the Farm Metric Server indicator are outlined
below.
The primary Farm Metric Server is active and has an accepted version of the
OK summary database software installed.
The summary database software version for the backup Farm Metric Server is
Warning not accepted because another server in the server farm has a later version of
the summary database software installed. You need to upgrade the software
on the backup Farm Metric Server. See Chapter 2, “Installing Resource
Manager” for details.
The summary database software version for the primary Farm Metric Server is
Critical not accepted because another server in the server farm has a later version of
the summary database software installed. You need to upgrade the software
on the primary Farm Metric Server. See Chapter 2, “Installing Resource
Manager” for details.
The meanings of the status icons for the Run-time indicator are outlined below.
The Database Connection Server is currently updating the database without

OK error.
A connection problem between the Database Connection Server and the

Critical DBMS that the summary database is on has occurred during a database
update. Ensure the Database Connection Server user identification and
password for the DBMS are correct. See Chapter 2, “Installing Resource
Manager” for details. Use the Server Log for the Database Connection Server
(click View Server Log) for further error information.
Not configured. In this state, no Resource Manager servers are creating and
Configured storing summary data for inclusion in the database.
Automatic database updates are temporarily stopped. This is known as Sleep

mode. See “Temporarily stopping database updates” in the Resource
Sleep Manager online help system for details.
The Database Connection Server is in an idle state between database

Snooze updates.
The meanings of the status icons for the Failed Import indicator are outlined below.

Not configured. No update was therefore attempted.
Configured
Data was successfully imported.
OK
An error has occurred: the import was not successful.
Critical
Managing the Resource Manager Summary Database

Use the following sections to help you manage how much hard drive space your
summary database will require over a period of time. The size of your summary
database, as it grows over time, depends on the following basic factors:
• The number of servers in the server farm
• The average number of processes run on a server each day
• The number of Resource Manager metrics you are storing in the database
• The average number of server events for a server each day
• The length of time the database records are kept
• The DBMS summary database transaction log
Note The transaction log maintains a history of the data transactions for the
summary database. Resource Manager does not automatically purge the DBMS
summary database transaction log. You need to configure how the DBMS
controls the transaction log to restrict its growth. See your DBMS
documentation for details.
Various categories of data are written to the database for the server or server farm.
These are:
• Process information
• Server metrics
• Session information
• Application metrics
• Server events (server-down/server-up)
Estimating Summary Database Growth

You can use the following equations to formulate an idea of how much information
will be stored in the summary database per day. Note that these estimates are very
approximate and each server or farm may vary considerably from the example. The
following schedule of data is for a Resource Manager farm of 100 servers under
typical loads.
Process information
• Estimate @ 600 sessions per day (with six processes per session)
• Estimate @ 140 bytes per row in process database table
Estimated total is 6 x 600 x 140 = 504,000 bytes per server per day
Server metrics
• Estimate @ 15 metrics per server (summarized at hourly intervals)
• Estimate @ 100 bytes per row in metrics database table
Estimated total is 15 x 100 x 24 = 36,000 bytes per server per day
Session information
• Estimate @ 100 sessions per server per day
• Estimate @ 100 bytes per row in session database table
Estimated total is 100 x 100 = 10,000 bytes per server per day
Sub Total is 84,000 + 36,000 + 10,000 = 130,000 bytes per server per day
Application metrics
• Estimate @ 20 application metrics per farm (summarized at hourly intervals)
• Estimate @ 100 bytes per row in application metric database table
Estimated total is 20 x 100 x 24 = 48,000 bytes for the farm per day
Server events
• Estimate @ 1% of farm servers restarted per day
• Estimate @ 20 bytes per row in event log database table
Estimated total is 1 x 20 = 20 bytes for the farm per day
Sub Total is 48,000 + 20 = 48,020 bytes for the farm per day
GRAND TOTAL is (504,000 + 36,000 + 10,000) x 100 + 48,020 = 52.50 MB of
summarized information stored in the database per day on a 100 Resource Manager
server farm. (Note: To obtain megabytes from bytes, we divided by 1024 squared.)
Managing Summary Database Growth

As described above for a typical farm, summary data stored in your database can be
substantial. Unless you manage your external database appropriately, it can grow
until all available storage space is used.
You need to monitor the rate at which your database is growing. Do this by
regularly checking how much disk space is available. You can then use this
information to help you set metrics and purge schedules for your database in order
to limit growth rates.
Some things you might want to do when using a summary database are to:
• Regularly check the available disk space on the database host computer so you
can work out an average for the amount of information being stored each day.
Tip You can configure your summary database DBMS to constrain database
size. See the Resource Manager online help system for instructions on how to
do this.
• Regularly create reports on the information you have stored and analyze which
metrics are appropriate for long-term storage and historical reporting for your
system.
If you find you are keeping metrics in the database unnecessarily, remove them
from the list being stored.
• Work out how often you need to check on resource usage for each metric.
You can create reports on these items on a regular basis and set up a purging
schedule to remove them from the database after you create the reports. The more
regular your reporting, the sooner you can purge the database of the information.
CHAPTER 5
Reporting and Analyzing Resource

Manager Information
Overview
This chapter describes the reports you can produce using data that is held on each
server or in a summary database. Topics include:
• Descriptions of the types of reports you can produce
• Generating detailed reports about current activity
• Generating summarized reports about past activity
• How reports from servers in different time zones and languages are handled
• Estimating the concurrent user capacity of a server
Detailed instructions on how to produce each type of report are provided in the
Resource Manager online help system.
You can also use the Report Center in the Access Suite Console to produce reports
from a variety of real-time and historical data sources. For further details of how to
use the Report Center, see the assistance provided on the Report Center screens, the
Access Suite Console online help system, or the MetaFrame Presentation Server
Administrator’s Guide.
Generating Reports to Analyze Data

Resource Manager enables you to produce two report types:
• Current: These reports are generated from Resource Manager information
stored in the local database on each server and can be generated either on a per
server basis or for multiple servers. Information is recorded at 15 second
intervals and is referred to as real-time.
• Summary: These reports are generated from data stored in the summary
database and can be generated either on a per-server basis or for multiple
servers. Summary reports are less detailed than current reports; however, they
can be generated for any times within the period for which data is stored in the
summary database. You can customize summary reports to include or exclude
various record types.
Important Summary reports include metric information only for times when
data was being stored in the summary database. If you set Resource Manager to
capture metric data only for certain times for a specific server, data from outside
these periods is not included.
All reports are displayed in a report viewer window. Reports contain navigation
links to allow you to step between the top of the report and any of the tables within
it. You can also print the report or save it in HTML or comma-separated values
(CSV) format.
Note Resource Manager uses a number of HTML templates to create reports.

These are located on each server in the Templates subfolder of the Resource
Manager folder. To avoid inconsistencies in the reports, do not edit the templates.
If your number format settings use commas as decimal separators, Resource
Manager replaces them with semicolons (;) when saving reports to CSV format
because commas are used specifically in this format to separate the items of data in
the file.
Creating Reports on Current Activity

Current reports can provide detailed information about the following:
• Statistics about current process activity or application usage in your server farm
• Statistics about current user activity in your server farm
• The status of a server at a particular moment
Go to Document Center Chapter 5 Reporting and Analyzing Resource Manager Information 57
Reporting on the Use of Processes or Applications

You can produce a report containing information about monitored processes in the
server farm or processes that are running on a specific server.
The report tells you about the resource usage of the process, the times it is active
and loaded, and the user(s) running it. For example, you might use this report to
examine a server's details when one or more of its metrics enter an alarm state.
The level of detail shown in the report depends on the options you select when you
set up the report. The report has a general information section at the top, including
the date and time of report generation, the report type, and the various report
options. Within the report, you can click links to navigate to the different sections.
The information shown in the report is as follows:
• Processes
The Processes table shows details about each instance of the selected process
(or all processes) being run on the server by the selected user (or all users) at the
time you generate the report. The table shows the following information:
• Name: The name of the process.
• Path: The location of the process on the server. This may help you
distinguish between instances.
• Product Date: The date from the program file. This gives an idea of when
the executable was created and may help you identify the process.
• Type: A Microsoft Windows-defined process type code.
• Version: The version number that is defined within the program file.
Note: This information may not be present for some programs.
• Start Time: The date and time at which this instance of the process was
loaded on the selected server. Times are shown in the local time zone of the
server being reported on.
• % Active: The length of time that this instance of the process was active, as
a percentage of the time since it was loaded on the server.
• User: The user name and domain of the user that is running this instance of
the process.
• CPU
The CPU table shows CPU usage statistics for the user(s) running the
process(es) at the time you generate the report. The table shows the following
information:
• Overall CPU Utilization

• Kernel mode
• User mode
• CPU Utilization while active
• Kernel mode
• User mode
For information about what these categories mean, consult the documentation
for the server operating system.
• Memory
The Memory table shows memory usage statistics for the user(s) running the
information:
• Working set while active
• Nominal working set
• Peak working set
• Peak paged pool
• Peak non-paged pool
• Peak page file usage
• Page faults/sec
• Times
The Times table shows process activity statistics for the user(s) running the
information:
• Time loaded: The length of time for which the processes were loaded on
the server.
• Time active: The length of time for which the processes were active.
• Active/loaded ratio: The length of time that the processes were active, as a
percentage of the time since they were loaded on the server.
• Total time loaded: The total length of time for which all the processes were
loaded.
• Total time active: The total length of time for which all the processes were
active.
• First/last recorded use: For all the processes that were running, the earliest
and latest times at which a process was started. Times are shown in the local
time zone of the server being reported on.
• Users
If you generate the report for all users, the Users table shows the user name and
domain of all the users running the processes, and the servers on which they are
running, at the time you generate the report. Remember, if you generated the
report for a single user, then the report includes only that user.
Reporting on User Activity

You can use Resource Manager to provide information about users who have active
sessions on a specific server at the time you generate the report. The user report tells
you about the users' sessions and the processes they are running.
Sessions and processes are shown as different sections in the report and are
described below. The report has a general information section at the top, including
the date and time of report generation, the report type, and the various selections
made when the report was set up. Within the report, you can click links to navigate
to the different sections.
The information shown in the report is as follows:
• Sessions
The Sessions table lists all the sessions that are being run by the selected user(s)
on the selected server at the time you generate the report. The table shows the
following information:
• User: The ID and domain of the user who is running the session.
• Session Name: The name of the session.
• Protocol: The protocol that was used for the session: for example, TCP/IP.
• Start Time: The date and time at which the session was started. Times are
shown in the local time zone of the server being reported on.
• Duration: How long the session has been running.
• Process Count: The number of processes that are running in the session.
This information is expanded in the Processes table further down the report.
• Processes
The Processes table shows a complete list of all the processes, and the number
of instances of each, run by the selected users on the selected server. The
processes are listed in order of start time, giving you a picture of what the user
was doing most recently.
Looking Back to a Specific Time on a Server

If there is a problem on a particular server, you can produce a Server Snapshot
report, showing its status at the time the problem occurred. You can then use this
report to evaluate why the problem happened.
The report tells you about users and processes utilizing the server and presents
information about monitored metrics.You can retrieve information for anytime
within the previous 96 hours.
You can also generate Server Snapshot reports from the real-time graph of a server
metric. See the Resource Manager online help system for more information.
Within the report, you can click links to navigate to the different sections. The
information shown in the report is as follows:
• Processes over 5% CPU or Memory
This table shows any processes that use more than 5% of the CPU load or
memory.
• Users and Processes
This table shows the users and the processes they ran on the selected server. The
table shows the following information:
• User: The user name and domain of each user who had an active session on
the server.
• Process: The name of each process that the user ran on the server. There is a
separate entry for each instance of the process.
• Path: The location of each process instance on the server. This may help
you distinguish between instances.
• Version: For each instance, the version number that is defined within the
program file. This information may not be present for some programs.
• Metrics
The Metrics table gives information about every metric that was monitored on
the server over the 15 second period. The table shows the following
information:
• Object: This is the category being monitored. It is a physical or logical
system resource: for example, a computer's hard drive.
Note The Terminal Services object is returned from Windows Terminal

Server. The Inactive Sessions counter for this object includes idle sessions.
• Instance: An individual example of the object or a state it needs to reach to

be counted. For example, a computer may have more than one hard drive. In
this case, the instance would identify which drive is being examined.
• Counter: This is the specific aspect of the object being monitored. For
example, free disk space.
• Time: The date and time at which the operating system last recorded the
metric. Dates are shown in dd/mm/yyyy format; times are shown in the local
time zone of the server being reported on.
• Value: The value of the metric at the time of sampling. Some metrics are
average values.
Creating Reports on Past Activity

Summary reports can provide summarized information about the following:
• Statistics about past process activity in your server farm
• Statistics about past user activity in your server farm
• The status of a server for a specified one hour period
Important Generating more than two summary reports at a time will overburden
most systems.
You can also generate reports from the summary database using an external
package such as Crystal Reports. To help you do this, the database schema used by
the summary database is described in Appendix B. Citrix provides several Crystal
Reports templates that you can use. These templates are available in several
languages, and are available for download from
http://www.citrix.com/download/.
Specific Processor Usage

You can use Resource Manager to provide information about the resource usage of
a process, users who ran it, and on what servers it ran over a selected period.
set up the report. Remember, if you select a specific server and/or user, the report
includes only the information related to that server or user.
The options you choose depend on why you are generating the report. For example,
you might include the CPU and memory statistics because you want to get an
indication of the load that a program imposes on the server.
information shown in the report is as follows:
• CPU
For processes that are finished, the CPU table shows CPU usage statistics for
the user(s) who ran the selected process and the server(s) on which it ran. The
table shows the following information:
• Overall CPU Utilization
• Kernel mode
• User mode
• CPU Utilization while active
• Kernel mode
• User mode
• Memory
For processes that are finished, the memory table shows memory usage
statistics for the user(s) who ran the process and the server(s) on which it ran.
The table shows the following information:
• Working set while active
• Nominal working set
• Peak working set
• Peak paged pool
• Peak non-paged pool
• Peak page file usage
• Page faults/sec
• Times
The Times table shows process activity statistics for the user(s) who ran the
process and the server(s) on which it ran. The table shows the following
information:
• Time loaded: The length of time for which the process was loaded on the
server.
• Time active: The length of time that the process was active.
• Active/loaded ratio: The length of time that the process was active, as a
percentage of the length of time that it was loaded on the server.
• Total time loaded: The total length of time for which the process was
loaded, for all its different locations and versions.
• Total time active: The total length of time for which the process was active,
for all its different locations and versions.
• First/last recorded use: The earliest and latest times at which the process
was started.
Times are shown in the local time zone of the server requesting the report.
• Users
The Users table shows the user name and domain of all the user(s) who ran the
process and the server(s) on which it ran. Remember, if you generate the report
for a single user/server, the report includes only that user/server.
• Servers
The Servers table shows a list of all the servers on which the user executed the
selected process. If you generate the report for a single server, this section of the
report is omitted.
Looking Back at Specific Users’ Activities

You can use Resource Manager to provide information about a user's resource
usage, sessions, and the processes used and on what servers, over a selected period.
set up the report. For example, you might use this report when you want to examine
the activity of a specific user; for example, to analyze work patterns.
following information is shown in the report:
• Session Summary
The Session Summary table lists all the selected user's sessions over the
specified report period, on the selected server(s), and for the selected
process(es). The table shows the following information:
• Start Time: The date and time at which the session started. Times are
shown in the local time zone of the console requesting the report.
• End Time: The date and time at which the session finished. Times are
shown in the local time zone of the console requesting the report. This area
remains blank if the session was still running when the summary database
was last updated.
• Duration: The length of time for which the session ran. This area remains
blank if the session was still running when the summary database was last
updated.
• Server: The server on which the session ran.
• Client: The name of the client device. If the user employs a server instead of
a separate client device, the client device name is shown as “Console.”
• Published Application: If the user connected to any published applications,
this column shows the names of the applications.
• Winstation: The Winstation or Sessionname for the session.
• Protocol: The protocol used for the session; for example, ICA.
• Process Count: The total number of processes that were run during the
session.
• Processes: A list of the unique process names that were run during the
session. For example, if the Process Count is shown as 6 and only one
process name is appears in the Processes, that process ran six times. If you
select the Process Summary option under Report Options when you
generate the report, you will see more detailed information about each
process in a Process Summary table.
• Favorite Processes
The Favorite Processes table lists the top 10 processes run by the selected user,
on the selected server(s), and of the selected process(es). The processes are
listed in descending order of most frequent use. The table shows the following
information:
• Count: The number of times the user ran the process.
• Process: The name of the process.
• Time Loaded: The total length of time that the user ran the process.
• Time Active: The total length of time for which the process was actively
operating (as opposed to just being loaded).
• % Active: The total length of time for which the process was active, as a
percentage of the total time for which it was loaded.
• Last Use: The date and time at which the user most recently ran the process.
• Process Summary
The Process Summary table shows a complete list of all the processes (or the
selected process, if you selected a single process) run by the selected user,
within the selected report period, on the selected server(s). The processes are
listed in order of earliest first, giving a picture of what the user was doing at
specific times. The table shows the following information:
• Start Time: The date and time at which the user started the process. Times
are shown in the local time zone of the console requesting the report.
• End Time: The date and time at which the process terminated. Times are
shown in the local time zone of the console requesting the report. This area
remains blank if the session was still running when the summary database
was last updated.
• Server: The server on which the user ran the process.
• Process: The name of the process.
• Exit Code: The exit code returned for the process, indicating the status of
the process when it terminated. Consult the software vendor for more
information about the exit codes that are used for specific processes.
• % Active: The total length of time for which the process was active, as a
percentage of the total time for which it was loaded.
• Statistics
The Statistics table shows general statistics about sessions run by the user on
the selected server(s) and for the selected process(es). The table shows the
following information:
• Session duration: The duration of all completed sessions that the user ran.
The table shows the average, minimum, and maximum session duration, and
the total duration of all sessions.
• Session duration per day: The duration of all completed sessions that the
user ran per working day.
Note Working days are defined as Monday to Friday.
• Sessions per day: The number of sessions run per working day.
• Processes per day: The number of processes run per working day.
• Processes per session: The number of processes per session calculated from
the process count in the Session Summary table against all the sessions that
the user ran.
• Servers
The Servers table shows a list of all the servers on which the user executed one
or more processes and the number of processes the user ran. Remember, if you
generate the report for a single server, the report includes information only for
that server.
Activities on a Specific Server

Server Summary reports produce statistical information for a particular server for a
one hour period. You can use this report to:
• See users and associated process activity for a farm server, including the process
path and version.
• See the metrics saved to the external database for a server, including their
associated object along with the metric counter and mean value. The time the
metric was recorded and the value are also displayed.
Users and processes, and metrics are shown as two different sections in the report
and are described below. The report has a general information section at the top,
including the date and time of report generation, the report type, and the various
selections made when the report was set up. Within the report, you can click links to
navigate to the different sections.
The following information is shown in the report:
• Users and Processes
This table shows the user(s) and the processes they ran on the selected server
over the selected hour. The table shows the following information:
• User: The user name and domain of each user who had an active session on
the server.
• Process: The name of each process that was run on the server. There is a
separate entry for each instance of the process.
• Path: The location of each process instance on the server. This may help
you distinguish between instances.
• Version: For each instance, the version number that is defined within the
program file. Process versions are available for some Win32 processes, but
not for Win16 or DOS processes.
• Metrics
The Metrics table gives information about every metric that was monitored on
the server over the selected hour. The table shows the following information:
• Object: This is the category being monitored. It is a physical or logical
system resource: for example, a computer's hard drive.
Note The Terminal Services object is returned from Windows Terminal

Server. The Inactive Sessions counter for this object includes idle sessions.
• Counter: This is the specific aspect of the object being monitored; for
example, disk free space.
• Instance: An individual example of the object or a state it needs to reach to

be counted. For example, a computer may have more than one hard drive. In
this case, the instance would identify which drive is being examined.
• Time: The date and time at which the operating system last recorded the
metric. Dates are shown in dd/mm/yyyy format; times are shown in the local
time zone of the server.
• Value: The average value of the metric over the selected hour.
Displaying Reports from Servers in Different Time Zones or

Locales
This section describes how reports are displayed if you have servers in farms that
are located in different time zones and/or are localized to different languages.
Reporting Server Is in a Different Time Zone

Times and dates that you enter when generating reports are understood by the
system to be in the local time for the server on which you are reporting.
All times and dates shown in the report tables are in the local time zone for the
server being reported on, with the exception of User Summary and Process
Summary reports. User Summary and Process Summary reports show session times
in the local time zone of the server requesting the report.
Report generation times, in the report header, show the time the report was
generated in the requesting server’s local time, plus any UTC (Universal Time,
Coordinated) offset. Server Summary reports also show UTC offset for times in the
Metrics table.
An example scenario is a server farm with servers in various parts of the world.
This example server farm has:
• Resource Manager servers located in New York, United States (UTC - 5 hours)
• Resource Manager servers in Berlin, Germany (UTC + 1 hour)
The server farm administrator generates a Server Summary report for the last six
hours from a server in New York at 13:00 hours local time. The report shows a
generation time of “13:00-05:00” (New York time). An event that occurred two
hours previously (11:00 hours in New York) in Berlin is shown as 16:00 hours—the
local time the event occurred in Berlin.
What if a Reporting Server Uses a Different Language?

Resource Manager software supports a number of languages. In a server farm with
differing language versions of Resource Manager present, reports are localized to
the locale where the console requesting the report is. In the above example, the
information from Berlin (German locale) would be reported in American English
(United States locale).
Estimating the Concurrent User Capacity of a Server

You can use the summary database to estimate the concurrent user capacity of a
server. The summary database stores information concerning CPU and memory
usage for various processes running on MetaFrame Presentation Server.
To determine user capacity

1. Either
• Add the server to the published applications in an existing farm
—Or—
• Create a new farm and limit user access to approximately 20 users per
server.
2. Using the information in this guide, and the Resource Manager online help
system, configure and enable the summary database.
3. Ask your users to launch and use the published applications running on the
server you are testing. Ensure that users continue to use the server over a
suitable period of time, in order to create a record of resource usage that reflects
your users’ normal working practices.
4. Create a Crystal report that queries the following:
• Average CPU and memory usage for the specific processes being assessed,
per user
• Average CPU and memory usage for other processes associated with a user,
such as explorer.exe, ctfmon.exe, osa.exe, wfshell.exe, csrss.exe,
svchost.exe, and winlogon.exe
• A defined threshold; for example, no more than 90% CPU usage and/or no
more than 3GB of RAM used
• A calculation to extrapolate the number of users that can be divided into the
threshold given the resource usage above
In general, the longer the time users work on the server, the more accurate the data
averages that can be collected from the summary database.
CHAPTER 6
Billing Users for Resource Usage
Overview
Resource Manager enables you to produce Billing reports based on the information
stored in your summary database.
Billing reports use the resource usage data from the summary database and a fee
profile to calculate the charges for users of the server farm. You define fee profiles
to reflect different charging rates and currencies.
You can organize individual users and/or user groups into billable groups known as
cost centers. You can also bill individual domain users or user groups.
When Resource Manager generates a Billing report, it calculates the charges by
multiplying the resources used during the report period by the associated fee.
All reports are displayed in a report viewer window. You can save a report in
HTML or comma-separated values (CSV) format for later printing, viewing, or
inclusion in documents.
The detailed steps you need to carry out to complete the tasks outlined in this
chapter are provided in the Resource Manager online help system.
Creating a Fee Profile

Before you can produce Billing reports, configure at least one fee profile. If a
Billing report is generated without using a fee profile, no cost information is
provided.
In each fee profile you specify a currency and a list of rates to charge for resource
usage. Resources you can include in a fee profile are:
• Session time
• CPU time
• Memory
• Process loaded time

• Process active time
Organizing Users into Cost Centers for Billing

You can organize users into cost centers. For example, to charge different
departments within an organization, you create a cost center for each department.
Each cost center is linked to a fee profile: it is best to create at least one fee profile
before you create cost centers.
The Summary database users cost center is a predefined cost center that cannot be
removed, edited, or renamed. This cost center is defined as all users who have an
entry in the summary database.
Important Local user groups on farm servers that are included in cost centers can
be billed only if the server in question is currently running.
Producing Billing Reports

There are two types of Billing reports: Cost Center and Domain Users.
Cost Center Reports

This type of Billing report is used to charge cost centers for their use of certain
resources that are being monitored in the server farm. You need to have at least one
fee profile and one cost center to generate useful Cost Center Billing reports. The
report tells you about the use of various chargeable resources by the cost center(s)
over a selected period.
If there are users who are not members of any cost center, you can generate Billing
reports for them by billing against domain users. See “Domain User Reports” on
page 71 for details.
You can avoid billing of system processes by configuring Resource Manager to
ignore those processes. See the Resource Manager online help system for more
information about ignoring processes.
Go to Document Center Chapter 6 Billing Users for Resource Usage 71
Domain User Reports

This type of Billing report is used to charge individual users or user groups within
Microsoft Windows domains for their use of certain resources that are being
monitored in the server farm. You need to create at least one fee profile to generate
useful Domain User Billing reports. The report tells you about the use of various
chargeable resources by the users over a selected period.
Information Presentation in the Report

The columns shown in the report depend on the options you choose in the Report
Options dialog box. The report is laid out on a per cost center or domain user basis,
and is further broken down into the sessions used over the report period. Totals for
each column are displayed at the end of the report.
• User: The domain user(s) or cost center(s) for which the report is being
generated.
• Session Start: The date and time each session began during the report period.
Dates are shown in dd/mm/yyyy format; times are shown in the local time zone
of the console requesting the report, along with any UTC offset.
• Session Elapsed Time: The total length of all session times during the report
period. Unfinished sessions are billed for the time elapsed during the report
period. Report Totals for session elapsed time at the end of the report shows the
combined total of all user sessions, including unfinished sessions, during the
report period.
• Process: The names of the processes run by the user over the course of each
session.
• CPU Time Used: The CPU time used for each process over the course of each
session during the report period. Report Totals for CPU time used at the end of
the report shows the combined total of all CPU time used during the report
period.
• Memory Used: The memory used for each process (in megabyte-minutes) over
the course of each session. Report Totals for memory used at the end of the
report shows the combined total of all memory used during the report period.
• Process Loaded Time: The length of time each process was loaded during each
session. Report Totals for process loaded time at the end of the report shows
the combined total of all process loaded times during the report period.
• Process Active Time: The length of time each process was active during each
session. Report Totals for process active time at the end of the report shows the
combined total of all process active times during the report period.
• Cost: The amount to be charged for each process during each session. Report
Totals for cost at the end of the report shows the combined total of all charges
during the report period.
• Fee Profile: The fee schedule and currency for chargeable resources is listed at
the end of the report.
CHAPTER 7
Troubleshooting
Overview
This chapter covers some common questions that you may encounter when using
Resource Manager and offers possible solutions.
Unexpected Behavior
I set up a feature in Resource Manager, but it doesn't seem to be working. What could be
the problem?
Ensure the server you are monitoring is upgraded to Resource Manager for Citrix
MetaFrame Presentation Server 4.0, Enterprise Edition.
It is possible to use Resource Manager in an environment where some servers in the
server farm are upgraded and others are not. However, if you are monitoring a
server that is not upgraded, certain aspects of the user interface will not work for
that server, even if the server from which you are running the Presentation Server
Console is upgraded.
For a published application, Resource Manager shows the application count as zero,
even though some instances of the application are running. How can I see the correct
application count?
Check that you specified the full path (rather than just the application executable) in
the Properties dialog box for the published application.
The IMA service became unresponsive while I was adding Resource Manager metrics to
a server. Why?
If the IMA service stops working while you are adding metrics for monitoring on a
server, it may be because you tried to add a metric was already added. If there are
duplicate instances of a metric, Resource Manager cannot monitor either instance
of that metric.
Some metric values reported by Resource Manager are negative. What should I do?
Some values returned by the Windows Performance API and displayed by
Resource Manager are negative. You can ignore these values.
I connected to a Resource Manager server in a different time zone and saw some
apparent time discrepancies. What is happening?
All dates and times displayed in the Presentation Server Console are in the context
of the server's time zone and current local time. When you connect to a server using
a MetaFrame Presentation Server Client, the clock shown on your remote desktop
is in the client device’s time zone.
If you look in the local database, note that the times stored there are in UTC
(Coordinated Universal Time) and, therefore, cannot be directly compared with the
times displayed in the Console.
My Database Connection Server is not updating the summary database. Why?

Your Database Connection Server might not be able to connect to the summary
database DBMS. If your Database Connection Server cannot communicate with the
summary database during a database update, on the Summary Database tab, under
Status, the lower indicator changes to Critical .
Your summary data is not lost; it is stored locally until the problem is rectified.
Note Data that is delayed in reaching the summary database due to Database
Connection Server problems is subject to purge settings once in the summary
database. For example, if you cannot update the summary database for a week and
some of your stored data is set to be purged after five days, when the problems are
rectified and the data is stored in the summary database, it is purged at the next
purge time because it is already five days old.
To view any problems that have occurred

3. Click View Server Log. The server log shows if any problems have occurred.
Go to Document Center Chapter 7 Troubleshooting 75
To address communication problems between the Database Connection Server

and the summary database
• Ensure the user identification password for the Database Connection Server is
correct. See “Setting a System Data Source Name” on page 22 for details.
• If the connection problem still exists, the Database Connection Server may have
failed. This is also indicated by a Server Down status icon in the Resource
Manager Watcher window. More information about the fault is available from
that server’s Resource Manager server log.
I generated a Resource Manager summary report and the information I expected to find
was not there. Where is it?
When you create summary reports, all information for the report is derived from
records stored in the summary database. If you encounter problems with your
summary reports, these may be due to the following reasons:
• The Resource Manager metric you want to report on is not set to be stored in the
summary database.
Check that the metric in question is set to be stored in the external database for
the relevant server. See “Selecting Server and Application Metrics to Record in
the Database” on page 46.
• The report was created during the 24-hour period between database updates.
If you create several reports for the same information between the 24 hourly
automatic database update times, the information in the reports does not change.
If you need information on a server for periods between database updates, use
Resource Manager Current reports. See Chapter 4, “Reporting and Analyzing
Information.”
• The report was created after the information was purged from the summary
database.
You must create reports on information before it is purged from the summary
database. Verify that your purging schedule gives you time to create appropriate
reports. See “Removing Unwanted Information from the Database” on page 47
for details about database purging schedules.
• The name of a contributing Resource Manager server is changed.
Each server in the server farm is identified to the summary database by the
server’s network identification computer name. If you change a server’s name,
existing records referencing the old name remain in the database until purged
while new records for the new server name are created.
Note If you change a server’s name from x to y, and rename another server in
the server farm from z to x, new reports for server x are collated from data from
both new server x and old server x.
I am getting the error message: “The Farm Metric Server cannot be contacted.” What
can I do?
If you see this message:
• First, find out which servers are acting as the primary Farm Metric Server and
the backup Farm Metric Server. (See below.)
• Ensure that both the primary Farm Metric Server and the backup are
operational. If either server is down, restart it.
If either server is heavily loaded, the Farm Metric Server can take some time to
respond, so this error message probably indicates a time-out error. If you suspect
that this might be the case, wait for a few moments or change the Farm Metric
Server and the backup to servers that are more lightly loaded.
Locating Servers
How do I find out which server is currently acting as the Database Connection Server?
To identify the Database Connection Server
1. On any Resource Manager server, start the Presentation Server Console.
2. In the left pane of the Console, click Resource Manager.
3. In the right pane, click the Summary Database tab. The current Database
Connection Server is shown.
See the Resource Manager online help system for instructions about how to change
the Database Connection Server.
How do I find out which server is currently acting as the Farm Metric Server?
To identify the Farm Metric Server
1. On any Resource Manager server, start the Presentation Server Console.
2. In the left pane of the console, click Resource Manager.
3. In the right pane, click the Farm Metric Server tab. The current Farm Metric
Servers are listed.
Go to Document Center Chapter 7 Troubleshooting 77
Permissions
Certain Resource Manager permissions provide differing levels of access
depending on whether you are using the Presentation Server Console or the Access
Suite Console. The following outline the most common issues you may encounter.
Note For more information on how to work wth permissions, see the on-screen
instructions in the Presentation Server Console, or the MetaFrame Presentation
Server Administrator’s Guide.
Using the Presentation Server Console, I can see the Servers node, am able to select a
server and to view server health. However, using the Access Suite Console, I cannot
access server health information. How can I view this information from the Access Suite
Console?
To be able to view server health information from the Access Suite Console, you
need to set an extra permission. Permissions are set using the Presentation Server
Console. Select the administrator for whom you want to change permissions in the
Presentation Server Console, and enable the View Resource Management
Configuration and Alerts permission. This permission is located under Properties
> Permissions > Resource Manager > Resource Management.
You should now be able to expand server nodes and view all Resource Manager
information from the Access Suite Console.
Using the Access Suite Console, I cannot see server nodes within the farm, am unable
to view any Resource Manager information, and am not receiving alerts. I can see this
information from my Presentation Server Console. How do I view this information from
the Access Suite Console?
You need to set an extra permission to view this information from the Access Suite
Console. Permissions are set using the Presentation Server Console. Select the
administrator for whom you want to change permissions, and enable the View RM
Information and Alerts. This permission is located under Properties >
Permissions > Servers > Resource Manager. You need to do this for each server
you want to monitor.
You should now be able to expand server nodes and view all Resource Manager
alerts from the Access Suite Console.
APPENDIX A
Default Metric Set
Overview
This appendix describes the default set of metrics that is monitored by Resource
Manager for Microsoft Windows 2000 and Windows Server 2003 operating
systems. The explanations for each metric are based on the default configuration.
Alternative configurations may produce alerts under different circumstances from
those described here.
Default Set of Metrics

Data Store Connection Failure
Minutes
This is the number of minutes since the server last successfully connected to the
data store, informing you if communications between a server and the IMA data
store fail. This failure could be because:
• The IMA data store DBMS system is down. This could be due to failure or for
maintenance
• The network connection to the server with the IMA data store is down
• The server with the IMA data store is down
Logical Disk
Important Resource Manager Logical Disk metrics require that Windows Logical
Disk counters are enabled. You can determine whether or not they are enabled by
running the diskperf utility at the command line. For more information on the
Logical Disk performance counters, run the diskperf /? command.
% Disk Time
Gives an indication of how busy the disks are. The disk can become a bottleneck for
a number of reasons:
• The server has too little physical memory so is “thrashing.” If thrashing is
occurring, the pages/sec will also be high.
• A single user is running an application or process that makes extensive and
rapid use of the disk. You can investigate such a user by running Current
Process and Current User reports.
• Many users are performing large amounts of disk activity. The speed of the
disks may be the server’s bottleneck.
The metric % Disk Time is calculated using a number of factors and values above
100% are possible. If you see values of 100% disk time, the disk is in constant use.
Values greater than 100% may indicate that the disk is too slow for the number of
requests.
% Free Space
The server is running out of disk space. Several factors can cause this:
• A lack of remaining disk space after installing the operating system and
applications
• A large number of users have logged on (now or in the past) and their
configuration data, settings, and files are taking up too much space
• A rogue process or user is consuming a large amount of disk space
Memory
Available Bytes
Informs you if too much memory is being used. This could be because:
• Too many users are logged on.
• The applications that users are running are too memory hungry for the amount
of memory available on the server.
• Some user or process is using a large amount of memory. Running a Current
Process report may help you track this down.
Being short on memory could result in “thrashing.” The disk usage and paging
metrics may also change to a red alarm state.
Go to Document Center Appendix A Default Metric Set 81
Pages/sec
A large amount of paging indicates either:
• The system is low on physical memory and the disk is being used extensively as
virtual memory. This can be caused by too many users being logged on, too
many processes running, or a rogue process “stealing” virtual memory.
—Or—
• An active process or processes are making large and frequent memory accesses.
Too much paging degrades the performance of the server for all users logged on.
The Available Bytes, Disk, and % Processor Time metrics may also enter warning
or danger states when a large amount of paging occurs. Short bursts of heavy
paging are normal, but long periods of heavy paging seriously affect server
performance.
Network Interface
Bytes Total/sec
Gives a good indication of how much network activity this server is generating or
receiving. If this metric changes to yellow or red, the server is experiencing
unusually high network activity and may cause a network saturation. If too many
users are remotely logged on for the network card to support, this metric may
change to a warning or danger state. In this situation, the bottleneck could be the
network or server’s network card, which may decrease performance of users’
sessions.
Paging File
% Usage
A high page file usage usually indicates that the server’s page file size should be
extended. If the Memory: Pages/sec metric is also high, it is a good idea to add
more physical memory.
Processor
% Interrupt Time
The processor is spending a large amount of time responding to input and output
rather than user processing. A large value for interrupt time usually indicates a
hardware problem or a very busy server.
% Processor Time
A high processor time for a long period of time indicates that the processor is the
bottleneck of the server, too many users are logged on, or there is a rogue user or
process (use the Current Process report to investigate).
System
Context Switches/sec
A large number of threads and/or processes are competing for processor time.
Terminal Services
Active Sessions
A large number of users are logged on and running applications. The server may
begin running out of memory or processor time and performance for users may
deteriorate. Note that current Presentation Server Console sessions are listed as
“active.”
Inactive Sessions
There is a large number of disconnected ssessions that are taking virtual memory.
Remove some disconnected sessions or reduce the length of time for which
disconnected sessions can persist until they are automatically removed.
APPENDIX B
Summary Database Schema
Overview
This appendix describes the layout and organization of the Resource Manager
summary database schema. The summary database is a data warehouse made up of
historical data imported from each Resource Manager server in the server farm. The
database schema of the local database is de-normalized, whereas the data stored in
the summary database is extensively normalized to save storage space. It includes:
• A diagram of database entity relationships
• Descriptions of each database entity table
Database Entity Relationship Diagram

The diagram on the following page shows all the tables in the schema and the links
required to retrieve data. The main tables are named SDB_xxx and have a white
background to their title bar. Supporting (or look-up) tables have a gray background
to their title bar and are (mostly) named LU_xxx. Some supporting tables are used
more than once. This reduces the space required to store the data.
A diagram showing database entity relationships

Go to Document Center Appendix B Summary Database Schema 85
Database Entity Table Descriptions

The following data types are all described using Microsoft SQL Server. For type
mappings to other SQL databases, see the type mapping section at the end of this
appendix.
Application History
SDB_APPHISTORY
This table stores a history of published applications supported on each server.
SDB_HISTORY
PK_SDB_APPHISTID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FK_APPNAMEID int NOT Pointer to LU_APPNAME, application name
NULL
FK_SERVERID int NOT Pointer to LU_SERVER, server name
NULL
STARTTIME date NOT Time application was supported by this
time NULL server
ENDTIME date NULL Time application no longer supported by
time server
SERVERUTCBIAS int NOT Bias in minutes to be subtracted from
NULL EVENTTIME to find the event time in the
server's local time zone
Foreign Key(s)
• FK_APPNAMEID
• FK_ SERVERID
Unique
• FK_ SERVERID
• FK_ APPNAMEID
• STARTTIME
Support Tables Referenced
• LU_APPNAME
• LU_SERVER
Application Metrics
SDB_APPMETRICS
This table stores a summary of all application metrics in a server farm.
SDB_APPMETRICS
FK_APPNAMEID int NOT Pointer to LU_APPNAME, application
NULL name and type
FK_FARMNAMEID int NOT Pointer to LU_FARMNAME, farm name
NULL
FK_OBJECTID int NOT Pointer to LU_OBJECT, object name
NULL
APPMETRICUPDATETIME date NOT Timestamp of last application metric data
time NULL point in dataset (stored in UTC)
APPMETRICSAMPLEPERIOD int NOT Sample period of summary record in
NULL seconds
APPMETRICDATACOUNT int NOT Number of data points used to
NULL summarize this row
MINAPPMETRICVALUE float NOT Minimum application metric value
NULL
MAXAPPMETRICVALUE float NOT Maximum application metric value
NULL
MEANAPPMETRICVALUE float NOT Mean application metric value
NULL
STDDEVAPPMETRICVALUE float NOT Standard deviation of application metric
NULL values
Foreign Key(s)
• FK_APPNAMEID
• FK_FARMNAMEID
• FK_OBJECTID
Unique
• FK_APPNAMEID
• FK_FARMNAMEID
• FK_OBJECTID
• APPMETRICUPDATETIME
Additional Indexed Columns
APPMETRICUPDATETIME
• LU_APPNAME
• LU_FARMNAME
• LU_OBJECT
Client History
SDB_CLIENTHISTORY
SDB_CLIENTHISTORY
FK_SDB_SESSIONID int NOT Pointer into SDB_SESSION for
NULL sessionID
FK_CLIENT_ID int NOT Pointer into LU_CLIENT for client name
NULL and address
FK_CLIENTPROPERTIESID int NOT Pointer into LU_CLIENT_PROPERTIES
NULL for build number, version, client type
STARTTIME date NOT Time client connected to session
time NULL
ENDTIME date NOT Time client disconnected (or time IMA
time NULL service was stopped)
NULL time fields to adjust the time to the
USINGSG INT NOT Boolean indicating whether Secure
NULL Gateway was used
FK_LAUNCHERID INT NOT Pointer into LU_LAUNCHER table
NULL
Foreign Key(s)
• FK_SDB_SESSIONID
• FK_CLIENTID
• FK_CLIENTPROPERTIESID
• FK_LAUNCHERID
• SDB_SESSION
• LU_CLIENT
• LU_CLIENTPROPERTIES
• LU_LAUNCHER
Connection History
SDB_CONNECTIONHISTORY
This table stores connected and disconnected session counts at the time of any
session connect or disconnect event for each server in the farm.
SDB_CONNECTIONHISTORY
FK_SERVERID int NOT Pointer to LU_SERVER, RM server name
NULL
CONNECTED int NOT Count of connected sessions
NULL
DISCONNECTED int NOT Count of disconnected sessions
NULL
TIMESTAMP date NOT Bias in minutes to be subtracted from
time NULL TIMESTAMP to adjust the time to the server's
local time zone
Foreign Key(s)
• FK_ SERVERID
Unique
• FK_ SERVERID
• TIMESTAMP

• LU_ SERVER
Event Log
SDB_EVENTLOG
This table stores generic IMA service up and IMA service down events that occur
on a server farm.
SDB_EVENTLOG
EVENTCODE int NOT Generic event ID. SERVER_DOWN = 0,
NULL SERVER_UP = 1
FK_SERVERID int NOT Pointer to LU_SERVER, Resource Manager
NULL server name
EVENTTIME date NOT Timestamp of event occurrence (Date and
time NULL Time)
server’s local time zone
Foreign Key(s)
FK_SERVERID
EVENTTIME
LU_SERVER
Unique
• EVENTCODE
• FK_SERVERID
• EVENTTIME
Administrator Configurable Server Metrics

SDB_METRICS
This table stores all metrics imported from each Resource Manager server in the
server farm. The metric values are summarized to reduce data storage requirements.
SDB_METRICS
NULL server name
FK_METRICID int NOT Pointer to LU_METRIC, metric description
NULL
METRICUPDATETIME date NOT Timestamp of last metric data point in dataset
time NULL (stored in UTC)
METRICSAMPLEPERIOD int NOT Sample period of summary record in seconds
NULL
METRICDATACOUNT int NOT Number of data points used to summarize
NULL this row
MINMETRICVALUE float NOT Minimum metric value
NULL
MAXMETRICVALUE float NOT Maximum metric value
NULL
MEANMETRICVALUE float NOT Mean metric value
NULL
STDDEVMETRICVALUE float NOT Standard deviation of metric values
NULL
FK_SERVERINFID Int NULL Reference to folders and zone information
Primary Key (Unique)

FK_SERVERID, FK_METRICID, METRICUPDATETIME
Foreign Key(s)
• FK_METRICID
• FK_SERVERID

METRICUPDATETIME
• LU_METRIC
• LU_SERVER
• LU_SERVERINF
Processes
SDB_PROCESS
This table stores process data per user.
SDB_PROCESS
PK_SDB_PROCESSID int NOT Unique identifier for referential integrity
NULL (Primary Key)
NULL server name
FK_PROCESSID int NOT Pointer to LU_PROCESS
NULL
FK_USERID int NOT Pointer to LU_USER, user name
NULL
FK_CLIENTID int NOT Pointer to LU_CLIENT, client name
NULL
NULL and type of the application with which the
process is most closely associated. This is
determined by examining each process in the
process parenting hierarchy (starting with the
process itself and working upwards) and
comparing the process executable path with
that of all published applications. If a match is
found, FK_APPNAMEID reflects this; if no
match is found, FK_APPNAMEID is set to
reflect a blank application name.
FK_SDB_SESSIONID int NOT Pointer to SDB_SESSION, session data
NULL
PID int NOT Process identifier (from operating system)
NULL
SDB_PROCESS
EXITCODE int NOT The exit code returned by the executable
NULL when it completes. 259 means “Still
executing”
AFFINITY int NOT A mask indicating which processor(s) the
NULL process can use to execute itself
STARTTIME date NOT Time the process started executing
time NULL
ENDTIME date NULL Time the process completed execution - or
time the time the process statistics were last
updated when EXITCODE = 259
TOTALTIME float NOT End time - Start time (in milliseconds)
NULL
ACTIVETIME float NOT A summation of all monitored periods of a
NULL process where the CPU time was greater
than 1%
KERNELUSE float NOT The percentage of kernel CPU time the
NULL process has used during its lifetime
USERUSE float NOT The percentage of user CPU time the
NULL process has used during its lifetime
USERACTIVE float NOT The percentage of user CPU that was being
NULL used during the ACTIVETIME
KERNELACTIVE float NOT The percentage of kernel CPU that was being
NULL used during the ACTIVETIME
MEMORY float NOT Sum of the average number of megabytes
NULL per minute used by the process during its
lifetime
MEMORYACTIVE float NOT The number of megabytes per minute used
NULL by the process during the ACTIVETIME
WORKSET int NOT The peak recorded working set of the
NULL processes at any point during its lifetime
PAGEFILE int NOT The peak page file allocated to process in
NULL bytes at any point in its lifetime
PAGEFAULTS int NOT The number of page faults that occurred
NULL
PAGEDPOOL int NOT The peak paged pool usage in bytes at any
NULL point in its lifetime
SDB_PROCESS
NONPAGEDPOOL int NOT The peak non-paged pool usage in bytes at
NULL any point in its lifetime
SESSID int NOT Matches the SESSIONID value in the
NULL SESSION table
SERVERUTCBIAS int NOT Bias, in minutes, to be subtracted from

• PK_SDB_PROCESSID
Foreign Key(s)
• FK_APPNAMEID
• FK_CLIENTID
• FK_PROCESSID
• FK_SERVERID
• FK_USERID
STARTTIME, ENDTIME
Unique
FK_SERVERID, STARTTIME, PID
• LU_APPNAME
• LU_CLIENT
• LU_PROCESS
• LU_SERVER
• LU_USER
• SDB_SESSION
User Information
SDB_SESSION
This table stores session data per user.
SDB_SESSION
PK_SDB_SESSIONID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FK_USERID int NOT Pointer to LU_USER, user name
NULL
NULL server name
FK_CLIENTID int NOT Pointer to LU_CLIENT, client name
NULL
NULL and type for the published application with
which the session was launched
FK_WINSTATIONID int NULL Pointer to LU_WINSTATION, name of the
WinStation through which the session is
connected
PROTOCOL int NULL
SESSIONSTART date NOT Start time of the first process run under the
time NULL user's session (stored in UTC)
SESSIONEND date NULL End time of the final process to end as part of
time the session (stored in UTC)
DURATION float NULL End time - Start time in milliseconds
SESSIONUTCBIAS int NOT Bias in minutes to be subtracted from
NULL SESSIONSTART in which to find the session
start time
SESSIONID int NOT A Session ID (generated by operating
NULL system)
TOTALTIMESUM float NOT Sum of the total time for all completed
NULL processes in milliseconds
ACTIVETIMESUM float NOT Sum of the active time for all completed
SDB_SESSION
CPUTIMESUM float NOT Sum of the CPU time for all completed
MEMORYSUM float NOT Sum of the memory usage for all completed
NULL processes
BILLSTATUS int NOT 0 = This session is not billed
NULL

• PK_SDB_SESSIONID
Foreign Key(s)
• FK_APPNAMEID
• FK_CLIENTID
• FK_SERVERID
• FK_USERID
• FK_WINSTATIONID
STARTTIME, ENDTIME
Unique
FK_SERVERID, SESSIONSTART, SESSIONID
• LU_APPNAME
• LU_CLIENT
• LU_SERVER
• LU_USER
• LU_WINSTATION
Version Control
SCHEMAVERSION
This table stores the version of the summary database schema. The version number
is queried on connection by the Database Connection Server to determine if it and
the summary database schema are compatible.
SCHEMAVERSION
VERSION int NOT Version number of database schema
NULL
Support and Look-Up Tables

LU_APPNAME
Look-up table of published application names.
LU_APPNAME
PK_APPNAMEID int NOT Unique identifier for referential integrity
NULL (Primary Key)
APPNAME nvarchar NOT Application name
(256) NULL
APPTYPE int NOT Application type: 0 = MetaFrame
NULL Presentation Server published
application, 1 = Resource Manager
application

PK_APPNAMEID
Unique
APPNAME, APPTYPE
LU_CLIENT
Look-up table of client names.
LU_CLIENT
PK_CLIENTID int NOT Unique identifier for referential
NULL integrity (Primary Key)
CLIENTNAME nvarchar(32) NOT Client name
NULL
CLIENTADDRESSFAMILY int NOT Client address family: 2 = AF_INET,
NULL 6 = AF_IPX
CLIENTADDRESS nvarchar(20) NOT Client address
NULL

PK_CLIENTID
Unique
CLIENTNAME, CLIENTADDRESSFAMILY, CLIENTADDRESS
LU_CLIENTPROPERTIES
Look-up table of client build, version (if available), and a reference to the
LU_CLIENTTYPEMAPPINGS table.
LU_CLIENTPROPERTIES
PK_CLIENTPROPERTIES int NOT Unique identifier for referential
FK_CLIENTTYPEID int NOT Reference to
NULL LU_CLIENTTYPEMAPPINGS
BUILD int NOT Client build number

NULL
VERSION nvarchar(64) NOT Client version, or “ “ if not available
NULL

PK_CLIENTPROPERTIESID
Unique
• FK_CLIENTTYPEID
• BUILD
• VERSION
LU_CLIENTTYPEMAPPINGS
Maps protocol type and client product identifier onto a client type name. Known
mappings are populated when the schema is created. New ones are added as ICA n
or RDP n where n is the client product identifier.
LU_CLIENTTYPEMAPPINGS
PK_CLIENTTYPEID int NOT Unique identifier for referential
PROTOCOL_TYPE int NOT Protocol type. 0 = Console, 1 = ICA,
NULL 2 =RDP
CLIENT_TYPE int NOT Client product identifier

NULL
CLIENTTYPENAME nvarchar(32) NOT Semi-useful string (for example,

NULL ICA Win32)

PK_CLIENTTYPEID
Unique
• CLIENT_TYPE
• PROTOCOL_TYPE
LU_FARMNAME
Look-up table for server farm names. It is a support table for LU_SERVER.
LU_FARMNAME
PK_FARMNAMEID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FARMNAME nvarchar NOT Server farm name
(255) NULL

PK_FARMNAMEID
Unique
FARMNAME
LU_INSTANCE
Look-up table of instances; for example, “C:\.”
LU_INSTANCE
PK_INSTANCEID int NOT Unique identifier for referential integrity
NULL (Primary Key)
INSTANCE nvarchar Instance name
(128) NULL

PK_INSTANCEID
Unique
INSTANCE
LU_LAUNCHER
Look-up table of launchers, for example, Program Neighborhood Agent.
LU_LAUNCHER
PK_LAUNCHERID int NOT Unique identifier for referential
LAUNCHER nvarchar(128) NULL Launcher name

PK_LAUNCHERID
Unique
LAUNCHER
LU_METRIC
Look-up table of metric definitions. This table stores look-up keys for objects,
metric counters, and instances.
LU_METRIC
PK_METRICID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FK_OBJECTID int NOT Pointer to LU_OBJECT, object name
NULL
FK_METRICCOUNTERID int NOT Pointer to LU_METRICCOUNTER, metric
NULL counter name
FK_INSTANCEID int NOT Pointer to LU_INSTANCE, instance name
NULL

PK_METRICID
Foreign Key(s)
• FK_INSTANCEID
• FK_METRICCOUNTERID
• FK_OBJECTID
Unique
• PK_OBJECTID, FK_METRICCOUNTERID, FK_INSTANCEID
Support Table Referenced
• LU_INSTANCE
• LU_METRICCOUNTER
• LU_OBJECT
LU_METRICCOUNTER
Look-up table of metric counters; for example, “% Disk time.”
LU_METRICCOUNTER
PK_METRICCOUNTERID int NOT Unique identifier for referential integrity
NULL (Primary Key)
METRICCOUNTER nvarchar NOT Metric counter name
(128) NULL

PK_METRICCOUNTERID
Unique
METRICCOUNTER
LU_NETDOMAIN
Look-up table for network domain names. It is a support table for LU_SERVER
and LU_USER.
LU_NETDOMAIN
PK_NETDOMAINID int NOT Unique identifier for referential integrity
NULL (Primary Key)
NETDOMAIN nvar NOT Network domain name
char NULL
(32)

PK_NETDOMAINID
Unique
NETDOMAIN
LU_OBJECT
Look-up table of objects; for example, “Logical Disk.”
LU_OBJECT
PK_OBJECTID int NOT Unique identifier for referential integrity
NULL (Primary Key)
OBJECT nvar NOT Object name
char NULL
(128)

PK_OBJECTID
Unique
OBJECT
LU_PATH
Look-up table of application paths. This is a support table for LU_PROCESS.
LU_PATH
PK_PATHID int NOT Unique identifier for referential integrity
NULL (Primary Key)
PATH nvar NOT Path
char NULL
(260)

PK_PATHID
Unique
PATH
LU_PROCESS
Look-up table of process details.
LU_PROCESS
PK_PROCESSID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FK_PATHID int NOT Pointer to LU_PATH, path name
NULL
FK_PROCESSNAMEID int NOT Pointer to LU_PROCESSNAME, executable
NULL name of the file associated with the process
TYPE int NOT Type of the executable. Win32, POSIX etc. -1
NULL indexed, where -1 = system process, 0
means Win32 application
VERSION nvar NOT Hexadecimal version number of executable
char NULL
(24)
PRODUCTDATE date NOT Timestamp of executable (by originators)
time NULL

PK_PROCESSID
Foreign Key(s)
• FK_PATHID
• FK_PROCESSNAMEID
Unique
FK_PATHID, FK_PROCESSNAMEID, TYPE, VERSION, PRODUCTDATE
• LU_PATH
• LU_PROCESSNAME
LU_PROCESSNAME
Look-up table of process names. This is a support table for LU_PROCESS.
LU_PROCESSNAME
PK_PROCESSNAMEID int NOT Unique identifier for referential integrity
NULL (Primary Key)
PROCESSNAME nvar NOT Executable name of the file associated with
char NULL the process
(255)

PK_PROCESSNAMEID
Unique
PROCESSNAME
LU_SERVER
Look-up table for Resource Manager server instances.
LU_SERVER
PK_SERVERID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FK_SERVERNAMEID int NOT Pointer to LU_SERVERNAME, Resource
NULL Manager server name
FK_NETDOMAINID int NOT Pointer to LU_NETDOMAIN, network domain
NULL name
FK_FARMNAMEID int NOT Pointer to LU_FARMNAME, server farm
NULL name
UPDATETIME date NOT The last time summary data was written to
time NULL the summary database for this particular
server

PK_SERVERID
Foreign Key(s)
• FK_FARMNAMEID
• FK_NETDOMAINID
• FK_SERVERNAMEID
Unique
FK_SERVERNAMEID, FK_NETDOMAINID, FK_FARMNAMEID
• LU_FARMNAME
• LU_NETDOMAIN
• LU_SERVERNAME
LU_SERVERNAME
Look-up table for Resource Manager server names. It is a support table for
LU_SERVER.
LU_SERVERNAME
PK_SERVERNAMEID int NOT Unique identifier for referential integrity
NULL (Primary Key)
SERVERNAME nvar NOT Server name
char NULL
(32)

PK_SERVERNAMEID
Unique
SERVERNAME
LU_SERVERINF
Look-up table for Resource Manager server folder and zone information.
LU_SERVERNAME
PK_SERVERINFID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FOLDER nvar NOT Folder name
char NULL
(255)
ZONE nvar NOT Zone name
char NULL
(128)

PK_SERVERINFID
Unique
FK_SERVERINFID, FOLDER, ZONE
LU_USER
Look-up table of user instances.
LU_USER
PK_USERID int NOT Unique identifier for referential integrity
NULL (Primary Key)
FK_NETDOMAINID int NOT Pointer to LU_NETDOMAIN, network domain
NULL name
USERNAME nvar NOT User name
char NULL
(32)

PK_USERID
Foreign Key(s)
FK_NETDOMAINID
Unique
FK_NETDOMAINID, USERNAME
LU_NETDOMAIN
LU_WINSTATION
Look-up table for WinStation names.
LU_WINSTATION
PK_WINSTATIONID int NOT Unique identifier for referential integrity
NULL (Primary Key)
WINSTATION nvar NOT Name of the WinStation through which the
char NULL session is connected
(32)

PK_WINSTATIONID
Unique
WINSTATION
SDB_SCRATCH
A cross-reference table used to identify session records contained in Billing reports
so that the sessions can be updated to show they were billed.
SDB_SCRATCH
CMC_ID int NULL Identifies the Presentation Server Console
that generated the report
REPORT_ID int NULL Identifies the report in question
USER_ID NULL Identifies a user within the report
SCRATCH_DATE date NOT Time that this table row was created
time NULL
FK_SDB_SESSIONID int NOT SDB_SESSION.PK_SDB_SESSIONID
NULL session table primary key value
SESSION_START date NULL SDB_SESSION.SESSIONSTART value
time
SESSION_DURATION float NULL SDB_SESSION.DURATION value
SDB_SCRATCH
PROCESS_TOTAL_TIME float NULL SDB_SESSION.TOTALTIMESUM value
SERVER_UTC_BIAS int NULL SDB_SESSION.SERVERUTCBIAS value
SERVER_NAME nvar NULL The server on which the session ran
char
(32)

None
Foreign Key(s)
None
Unique
None
• SESSION_START
• CMC_ID, REPORT_ID, USER_ID, FK_SDB_SESSIONID
SDB_HEURISTICS
This table holds miscellaneous data used by Resource Manager.
SDB_HEURISTICS
PK_HEURISTIC nvar NOT Name of heuristic
char NULL
(64)
HEURVALUE float NOT Value of heuristic
NULL
SQL Data Type Mapping

Column data types vary between different SQL databases. A list of all data types
used in the summary database for all supported SQL databases follows:
Description Microsoft SQL Server Oracle

Integer numbers int int
Small integer numbers tinyint smallint
Floating point numbers float float
Date and time datetime date
National variable length strings nvarchar varchar2
(Unicode)
Glossary
CSV Comma-separated values. A file format used as a portable

representation of a database. Each line is one entry or record and the
data fields in a record are separated by commas. Commas can be
followed by spaces and/or tab characters that are ignored. If a field
includes a comma, the whole field must be surrounded with double
quotes.
custom An administrator who is subordinate to a full administrator. Custom

administrator administrators cannot set up other administrator accounts and have
only a subset of the permissions that a full administrator has.
Database A Resource Manager server that writes data to, and reads data from,
Connection Server a summary database.
data store A data store that centralizes configuration information about published
applications, users, printers, and servers. Each server farm has a
single data store.
data source name The system data source name (DSN) stores information about how a
client can connect to a database. It is used by a client to access a
database management system (DBMS). In the case of Resource
Manager, the client is the Database Connection Server.
DBMS Database management system. A software interface between the

database and the user. A DBMS handles user requests for database
actions with provision for data security and integrity requirements.
Farm Metric Server This server interprets farm-wide metrics and then processes them as
part of its summary data. Application count is an example of a farm-
wide metric. You can also have a backup Farm Metric Server.
fee profile Rates that are charged for using different types of resources. Fee
profiles are used for billing.
full administrator An administrator who has full access to all the administrative functions
and features of the server farm. Full administrators are the only
administrators who are allowed to create or modify other administrator
accounts.
Independent Citrix’s server-to-server infrastructure that provides robust, secure,

Management and scalable tools for managing a server farm of any size. Among
Architecture (IMA) other features, IMA enables centralized platform-independent
management, an ODBC-compliant data store, and a suite of
management products that plug in to the Presentation Server
Console.
metric One of a series of measurable items for a server or application. You

can select which metrics you want to monitor for a particular server.
monitoring This occurs when Resource Manager is actively looking at the data on
servers.
object A server or published application.
process An instance of a program that is being executed.
Resource Manager An application that is not a MetaFrame Presentation Server published

application application but is still recognized by the Resource Manager system.
schema A description of a database to a DBMS in the language provided by

the DBMS. A schema defines aspects of the database, such as
attributes (fields) and attribute parameters.
purge To systematically eliminate old or unneeded information.
server farm A group of servers that are managed as a single unit, and that share
some form of physical connection and a single data store.
Go to Document Center Appendix 113
status icon A colored signal in the status display that shows the status of each
metric. When a status icon in the display changes, an alarm condition
occurs.
summary data An averaged calculation of metrics information recorded on a
Resource Manager server once each hour. Summarized data is
stored by a DBMS for reporting purposes.
UTC Coordinated Universal Time. UTC is the same time as Greenwich
Mean Time (GMT) and is the reference time zone used for calculating
world time zones.
Go to Document Center 115
Index
A D
Acrobat Reader, requirements 10 data source name 21
alarms definition 111
automatic alerts for 37 setting 22
counter 31 setting for Microsoft SQL Server 22
instance 31 setting for Oracle 23
object 31 data store
pausing notification of 37 definition 111
watching for 36 Database Connection Server
alerts 37 configuring 24
configuring automatic alerts 37 definition 111
email 38 DBMS
MAPI email configuration 38 definition 111
pausing 37 installing onto a server 18, 21
SMS (cell/mobile phone) 40 requirements 18, 21
SMTP email 38 transaction log 52
SNMP 41 default metrics 79
when a server stops operating 36 data store connection failure 79
applications logical disk 79
reporting on use 57 memory 80
selecting Count metric for 46 network interface 81
automatic alerts 37 paging file 81
processor 81
B system 82
terminal services 82
billing default metrics set 79
cost centers 70 dialog box
domain users 71 Data Sources (ODBC) 22–23, 39
Microsoft SQL Server DSN Configuration 22
C Oracle ODBC Driver Configuration 24
Summary Database Configuration 24
cell/mobile phone (SMS) alerts 40
cost centers 70
counter, for metrics 31 E
CSV files email alerts 38
definition 111 configuring in Resource Manager 40
saving to 56, 69 creating a mail profile 39
custom administrator MAPI Connection Server 38
definition 111 Resource Manager Mail Service 39
errors
fixing user identification/password conflict 75
missing summary report information 75
viewing Resource Manager server log files 74
F Microsoft SQL Server DSN Configuration dialog box 22

modems (for SMS alerts) 41
Farm Metric Server monitoring
configuration 49 definition 112
contacting 76 entire server farm 28
definition 111 real-time 31
fee profile 69 single server 29, 36
full administrator
definition 112
N
G negative metric values 74
network interface metrics 81
gateways for SMS alerts 41 Bytes Total/sec 81
new features 13
I
icons, meaning of 33 O
Independent Management Architecture, definition 112 object
installation definition 112
changing location after installation 30 object, for metrics 31
software requirements 17 Oracle ODBC Driver Configuration dialog box 24
instance, for metrics 31
P
L
paging file metrics 81
licensing 18 % Usage 81
log file, server 33 pausing alarms and alerts 37
logical disk metrics 79 performance monitor 42
% Disk Time 80 process
% Free Space 80 definition 112
Process Summary report 61
M processes
reporting on current 57
MAPI email
reporting on history of 61
to configure 38
processor metrics 81
memory metrics 80
% Interrupt Time 81
Available Bytes 80
% Processor Time 82
Pages/sec 81
profile for email alerts 39
metric
purge
checking status of 36
definition 112
configuring automatic alerts for 37
purging, summary database
counter 31
overview 47
Count, selecting applications for 46
default set of 79
definition 112 R
instance 31 real-time monitoring 31
issues to consider when selecting for summary
database 45
negative values 74
object 31
selecting for summary database 46
Go to Document Center Index 117
reports summary database

about current activity 56 DBMS supported 18
about past activity 61 estimating size 54
Billing, overview 69 growth management 54
Current Process 57 ignoring specific times during the day 47
Current User 59 issues to consider when selecting metrics for 45
overview 56 purging 47
Process Summary 61 size considerations 52
saving 67 transaction log 52
saving Billing 72 turning off 26
Server Snapshot 60 turning on 25
Server Summary 66 Summary Database Configuration dialog box 24
templates 56 summary database schema
User Summary 63 administrator configurable server metrics 90
viewing saved Billing 72 application metrics 85
viewing saved current and summary 67 LU_APPNAME 96
Resource Manager application LU_CLIENT 97
definition 112 LU_FARMNAME 97
Resource Manager Mail Service 39 LU_INSTANCE 99
Resource Manager server log, viewing 74 LU_LAUNCHER 99
Resource Manager tab 28 LU_METRICCOUNTER 101
LU_NETDOMAIN 101
S LU_OBJECT 102
LU_PATH 102
schema LU_PROCESS 103
definition 112 LU_PROCESSNAME 104
server LU_SERVER 104
determining user capacity 68 LU_SERVERINF 106
monitoring single 36 LU_SERVERNAME 105
reporting on recent activity of 60 LU_USER 106
server farm LU_WINSTATION 107
definition 112 processes 91
server log 33 SCHEMAVERSION 96
Server Summary report 66 SDB_APPMETRICS 86
SMS alerts 40 SDB_CLIENTHISTORY 87
configuring 41 SDB_EVENTLOG 89
modems 41 SDB_METRICS 90
SMTP email alerts SDB_PROCESS 91
configuring 37 SDB_SCRATCH 107
SNMP alerts 41 SDB_SESSION 94
to configure 41 SQL data type mapping 109
status icon support and look-up tables 96
definition 113 user information 94
status icons version control 96
meaning of 33 Summary Database tab 24
summary data suspending alarms and alerts 37
definition 113 system metrics 82
ignoring specific times during the day 47 Context Switches/sec 82
system requirements
DBMS 18, 21
for email alerts 38
for SMS alerts 40
for SNMP alerts 41
T
tab
Resource Manager 28
Summary Database 24
TAPI Servers 41
terminal services metrics 82
Active Sessions 82
Inactive Sessions 82
traps
List of SNMP 41
troubleshooting 73
U
unexpected behavior 73
uninstalling Resource Manager 20
User Summary report 63
users
reporting on current 59
reporting on history of 63
UTC
definition 113

Resource Manager Guide

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Resource Manager Guide

Transféré par

Droits d'auteur :

Formats disponibles

Resource Manager Administrator’s Guide

For other guides in this document set, go to the Document Center

Resource Manager for MetaFrame® Presentation Server

Chapter 2 Installing Resource Manager

Displaying Resource Manager and its Components . . . . . . . . . . . . . . . . . . . . . . . . 27

Chapter 3 Monitoring Servers and Applications in Real Time

Chapter 4 Recording the History of Servers and Applications

Scheduling Summary Data Collection and Removal . . . . . . . . . . . . . . . . . . . . . . . 47

Chapter 5 Reporting and Analyzing Resource Manager Information

Chapter 6 Billing Users for Resource Usage

Appendix A Default Metric Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79

Appendix B Summary Database Schema . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83

About this Guide

Introducing Resource Manager

Monitoring Your Existing Server Farm

Identifying, Diagnosing, and Solving Problems

Gauging and Justifying Future Resource Needs

Planning and Scaling Your Server Farm

Billing Users for Resource Usage

Receiving Failed Import Alerts

Receiving License Server Connection Failure Alerts

Resource Manager and the Access Suite Console

Extended Real-Time Monitoring Capabilities

What to Read Next

• Detailed information on Resource Manager-related tasks that you carry out

Installing Resource Manager

Installing Resource Manager

Selecting a Server to Use as the Farm Metric Server

If You Are Upgrading from a Previous Version

Installing or Upgrading Resource Manager

To install Resource Manager on a server

Uninstalling Resource Manager

To uninstall Resource Manager from a server

Setting Up a Summary Database

2. Set a system data source name (DSN).

Setting a System Data Source Name

Important You must type rmsummarydatabase exactly. Any spaces or

Screenshot of the Microsoft SQL Server DSN Configuration dialog box

To set a system Data Source Name for Oracle DBMS

1. Choose a server to be your Database Connection Server.

Important You must enter rmsummarydatabase exactly. Any spaces or

Configuring a Database Connection Server

To configure a Database Connection Server

Note Resource Manager supports Windows NT authentication for the

6. Click Test to check the connection to the database.

Turning the Summary Database On

To turn the summary database on

Turning the Summary Database Off

To turn the summary database off

Displaying Resource Manager and its Components

To open the Presentation Server Console

To install the Presentation Server Console on a workstation

Displaying the Main Resource Manager Screen

Screenshot of the main Resource Manager screen

Displaying Resource Manager for the Entire Server Farm

To display Resource Manager for the entire server farm

Displaying Resource Manager for a Single Server

Displaying the Applications in a Server Farm

Changing the Location of Resource Manager after Installation

To change the location of Resource Manager

Monitoring Servers and

Configuring Metrics on Servers