Académique Documents
Professionnel Documents
Culture Documents
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 7 Troubleshooting
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Unexpected Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Locating Servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Permissions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
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
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
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:
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.
Go to Document Center Chapter 1 Introduction 13
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.
Go to Document Center Chapter 1 Introduction 15
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.
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
Go to Document Center
CHAPTER 2
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.
18 Resource Manager Administrator’s Guide Go to Document Center
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.
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.
20 Resource Manager Administrator’s Guide Go to Document Center
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.
Go to Document Center Chapter 2 Installing Resource Manager 21
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.
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.
Note The system DSN setup described below may be different for other Oracle
versions.
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).
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.
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 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.
26 Resource Manager Administrator’s Guide Go to Document Center
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.
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.
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.
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.
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.
Go to Document Center
CHAPTER 3
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.
32 Resource Manager Administrator’s Guide Go to Document Center
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.
Represents normal operation; that is, the value of the metric falls between the
OK set limits.
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.
34 Resource Manager Administrator’s Guide Go to Document Center
Note You cannot type directly in the box, you must click Browse.
Go to Document Center Chapter 3 Monitoring Servers and Applications in Real Time 35
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.
Alerts are generated whenever the IMA service stops operating, including expected
events such as scheduled reboots.
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.
38 Resource Manager Administrator’s Guide Go to Document Center
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.
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.
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.
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.
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
An application example
for the purposes
of demonstrating the
having a Summary
Database.
Database
Server
An Application
An application example
for the purposes
of demonstrating the
having a Summary
Database.
Server
Server
Legend
An Application
Farm-wide metrics
An application example
for the purposes
of demonstrating the
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
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.
Note Collection restrictions may affect resource billing because process data that
is not stored in the summary database cannot be billed.
Go to Document Center Chapter 4 Recording the History of Servers and Applications 47
• 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)
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.
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.
The meanings of the status icons for the Database Connection Server
On/Off/Configuration indicator are outlined below.
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.
Note: This situation can arise only when there are mixed releases of the
summary database software within a server farm.
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.
Note: This situation can arise only when there are mixed releases of the
summary database software within a server farm.
Go to Document Center Chapter 4 Recording the History of Servers and Applications 51
The meanings of the status icons for the Run-time indicator are outlined below.
The meanings of the status icons for the Failed Import indicator are outlined below.
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)
Go to Document Center Chapter 4 Recording the History of Servers and Applications 53
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.
Go to Document Center
CHAPTER 5
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.
56 Resource Manager Administrator’s Guide Go to Document Center
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.
• 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.
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/.
Within the report, you can click links to navigate to the different sections. The
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
For information about what these categories mean, consult the documentation
for the server operating system.
• 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
For information about what these categories mean, consult the documentation
for the server operating system.
• 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.
Go to Document Center Chapter 5 Reporting and Analyzing Resource Manager Information 63
• 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.
• 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:
Go to Document Center Chapter 5 Reporting and Analyzing Resource Manager Information 65
• 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.
• 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.
66 Resource Manager Administrator’s Guide Go to Document Center
• Counter: This is the specific aspect of the object being monitored; for
example, disk free space.
Go to Document Center Chapter 5 Reporting and Analyzing Resource Manager Information 67
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.
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.
• 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.
Go to Document Center
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.
74 Resource Manager Administrator’s Guide Go to Document Center
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.
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.
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.
76 Resource Manager Administrator’s Guide Go to Document Center
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.
Go to Document Center
APPENDIX A
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.
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.
80 Resource Manager Administrator’s Guide Go to Document Center
% 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.
82 Resource Manager Administrator’s Guide Go to Document Center
% 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.
Go to Document Center
APPENDIX B
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
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
86 Resource Manager Administrator’s Guide Go to Document Center
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
Go to Document Center Appendix B Summary Database Schema 87
Unique
• FK_APPNAMEID
• FK_FARMNAMEID
• FK_OBJECTID
• APPMETRICUPDATETIME
Additional Indexed Columns
APPMETRICUPDATETIME
Support Tables Referenced
• 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)
SERVERUTCBIAS int NOT Bias in minutes to be subtracted from
NULL time fields to adjust the time to the
server's local time zone
USINGSG INT NOT Boolean indicating whether Secure
NULL Gateway was used
FK_LAUNCHERID INT NOT Pointer into LU_LAUNCHER table
NULL
88 Resource Manager Administrator’s Guide Go to Document Center
Foreign Key(s)
• FK_SDB_SESSIONID
• FK_CLIENTID
• FK_CLIENTPROPERTIESID
• FK_LAUNCHERID
Support Tables Referenced
• 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
Go to Document Center Appendix B Summary Database Schema 89
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)
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_SERVERID
Additional Indexed Columns
EVENTTIME
Support Tables Referenced
LU_SERVER
Unique
• EVENTCODE
• FK_SERVERID
• EVENTTIME
90 Resource Manager Administrator’s Guide Go to Document Center
SDB_METRICS
FK_SERVERID int NOT Pointer to LU_SERVER, Resource Manager
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)
SERVERUTCBIAS int NOT Bias in minutes to be subtracted from
NULL EVENTTIME to find the event time in the
server's local time zone
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
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)
FK_SERVERID int NOT Pointer to LU_SERVER, Resource Manager
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
FK_APPNAMEID int NOT Pointer to LU_APPNAME, application name
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
92 Resource Manager Administrator’s Guide Go to Document Center
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
Go to Document Center Appendix B Summary Database Schema 93
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
NULL EVENTTIME to find the event time in the
server's local time zone
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
FK_SERVERID int NOT Pointer to LU_SERVER, Resource Manager
NULL server name
FK_CLIENTID int NOT Pointer to LU_CLIENT, client name
NULL
FK_APPNAMEID int NOT Pointer to LU_APPNAME, application name
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
SERVERUTCBIAS int NOT Bias in minutes to be subtracted from
NULL EVENTTIME to find the event time in the
server's local time zone
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
NULL processes in milliseconds
Go to Document Center Appendix B Summary Database Schema 95
SDB_SESSION
CPUTIMESUM float NOT Sum of the CPU time for all completed
NULL processes in milliseconds
MEMORYSUM float NOT Sum of the memory usage for all completed
NULL processes
BILLSTATUS int NOT 0 = This session is not billed
NULL
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
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
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
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
NULL integrity (Primary Key)
FK_CLIENTTYPEID int NOT Reference to
NULL LU_CLIENTTYPEMAPPINGS
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
NULL integrity (Primary Key)
PROTOCOL_TYPE int NOT Protocol type. 0 = Console, 1 = ICA,
NULL 2 =RDP
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
Go to Document Center Appendix B Summary Database Schema 99
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
LU_LAUNCHER
Look-up table of launchers, for example, Program Neighborhood Agent.
LU_LAUNCHER
PK_LAUNCHERID int NOT Unique identifier for referential
NULL integrity (Primary Key)
LAUNCHER nvarchar(128) NULL Launcher name
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
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
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)
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)
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)
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
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)
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
Foreign Key(s)
• FK_FARMNAMEID
• FK_NETDOMAINID
• FK_SERVERNAMEID
Unique
FK_SERVERNAMEID, FK_NETDOMAINID, FK_FARMNAMEID
Support Tables Referenced
• 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)
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)
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)
Unique
FK_NETDOMAINID, USERNAME
Support Tables Referenced
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)
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
108 Resource Manager Administrator’s Guide Go to Document Center
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)
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
Go to Document Center Appendix B Summary Database Schema 109
Glossary
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.
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.
112 Resource Manager Administrator’s Guide Go to Document Center
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.
monitoring This occurs when Resource Manager is actively looking at the data on
servers.
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
116 Resource Manager Administrator’s Guide Go to Document Center
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