Working with Performance Monitor Charts and Analytics

This chapter provides an overview of Performance Monitor charts and analytics and discusses how to:

View user request analytics.
View component trace information.
View component statistics.
View portal statistics.
View PIA statistics.
View top portal content requests.
View top PeopleCode events.
View top PeopleCode executions.
View top PeopleCode SQL statements.
View top components.

Performance Monitor Charts and Analytics

This section contains an overview of Performance Monitor charts and analytics, list common elements, and lists the pages used to view user request analytics.

Performance Monitor Charts and Analytics Overview

Performance Monitor charts enable you to view charts and graphs that help you analyze and communicate performance information.

This information is derived from completed PMU data that is stored in the historic performance data tables.

Note. This interface is not designed to assist you in diagnosing real-time, current user issues, such as a stalled user request.

Note. The Standard Deviation (Std. Dev.) is provided on many pages. The standard deviation is a statistic that tells you how tightly all the values that are used to compute the average are clustered around the average. Large standard deviations warn that the averages appearing in the chart are not a reliable indicator of response times experienced by individual users.

Common Elements Used in This Chapter

Chart Legends	Many of the charts use color to differentiate performance data. Always use the labels and the legend to make sure that you interpret the charts correctly.
View in Grid	Enables you to view chart information in a grid format for sorting and downloading to Microsoft Excel.

Pages Used to View Performance Monitor Charts

Page Name	Object Name	Navigation	Usage
User Requests	PSPMTRANSUSERSUM	PeopleTools, Performance Monitor, Analytics, User Requests	This page shows top-level, completed PMUs that were initiated by a specific user ID. A user request is a set of PMUs that together make up a complete round-trip to the server. You can click the bars in the chart to drill down into the details of a request (User Request Details page).
Component Trace	PSPMCOMPPERF	PeopleTools, Performance Monitor, Analytics, Component Trace	Only used in conjunction with a performance trace. This page enables you to isolate the component performance that is associated with a particular performance trace.
Component Statistics	PSPMCOMPSTAT	PeopleTools, Performance Monitor, Analytics, Component Statistics	Displays performance metrics that are related to the SQL, PeopleCode, and buffer sizes of PeopleSoft components. In addition, the page displays standard deviations and average charts.
Portal Statistics	PSPMPORTALSTATS	PeopleTools, Performance Monitor, Analytics, Portal Statistics	Displays portal activity duration per user ID within a specified interval.
PIA Statistics	PSPMPIASTATS	PeopleTools, Performance Monitor, Analytics, PIA Statistics	Displays PIA activity duration and response size per user ID within a specified interval.
Top Portal Content Requests	PSPMTOPPORTAL	PeopleTools, Performance Monitor, Analytics, Top Portal Content Requests	Enables you to identify candidates for performance tuning by displaying the most requested content and the associated performance measurements.
Top PeopleCode Events	PSPMPCODEEVENTS	PeopleTools, Performance Monitor, Analytics, Top PeopleCode Events	Enables you to identify candidates for performance tuning by examining the top execution count, cumulative durations, and average durations of PeopleCode events.
Top PeopleCode Executions	PSPMTOPPCODE	PeopleTools, Performance Monitor, Analytics, Top PeopleCode Executions	Enables you to identify candidates for performance tuning by examining the top execution count, cumulative durations, and average durations of PeopleCode executions.
Top SQL Statements	PSPMSQLEVENTS	PeopleTools, Performance Monitor, Analytics, Top PeopleCode SQL Statements	Enables you to identify candidates for performance tuning by examining the top execution count, cumulative durations, and average durations of SQL statements that are executed by PeopleCode programs.
Top Components	PSPMCOMPEVENTS	PeopleTools, Performance Monitor, Analytics, Top Components	Enables you to identify candidates for performance tuning by examining the top execution count, cumulative durations and average durations of component actions.

Viewing User Request Analytics

Access the User Requests page.

This page shows top-level, completed PMUs that were initiated by a specific user ID. A top-level PMU is a PMU with no parent PMU. A user request is a set of PMUs that together make up a complete request to the monitored system.

Note. All rows in the User Request chart appear in ascending order by agent start date and time. Read the chart from the bottom to the top. If the date is the same between two adjacent rows, the system hides the date and shows only the time. If two rows have identical dates and times, the system adds a number (#x) to the label to differentiate the two bars. The charting tool requires the labels of the bars to be distinct.

You must enter either the user ID or the performance trace name as part of the search criteria.

System ID	Identifies each monitored system. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the lookup program (PSPM_LOOKUP). When searching on a user ID, generally, the system attempts to match a value that is stored in the OPRID (user ID) field of the PMU history table. However, because some top-level PMUs have "_unknown_" as the OPRID, the system attempts a second match. If the top-level PMU contains "_unknown_" as the OPRID, the system searches for a PMU 113 (Authenticate) in the same PMU set (same top instance ID) where PMU 113 contains the given user in the OPRID field.
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Static Scale	Enables you to choose the scale of the X-axis in the User Requests chart that is generated from the search results. The search results appear in result sets, which are groups of 10. Select this option to view a static scale. Note. After you select this option, you can't clear Static Scale until you refresh the page from the browser or reenter the page. The system behavior for each type of scale is: Dynamic scale. The system scales the X-axis according to the longest duration in the current result set. Static scale. A static scale displays the same scale for the X-axis across all of the result sets. A static scale is based on the largest duration, or bar, that exists among all of the result sets.
Search	The Search button performs a search based on the search criteria that is defined on the page. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Refresh	Updates the To Time value to equal the current time before performing a search. This enables you to view the most current results of a search.

The User Requests Chart

The User Requests chart reveals the time that a user initiated a specific request and its duration. The system displays user requests in sets of 10.

A single user request is represented by a bar within the bar chart. You can click any bar in the graph to view the associated User Request Detail component, which enables you to drill down into the details of a user request.

On this chart, the View in Grid hyperlink displays the chart information in a grid for downloading and customized sorting. The grid contains a hyperlink that accesses the User Request Details page.

The rows on the chart contain pop-up messages that appear when you move the cursor over the bar. The messages are:

For PMU 100 (Portal Request), the message reads Portal: [action] Duration (secs): yyy.
For PMU 101 (PIA Request), the message reads PIA: [action] Duration (secs): yyy.
For PMU 102 (Compression Request), the message reads Compression Request: [content type] Duration (secs): yyy.
All other PMUs show the PMU label followed by the duration.

See User Request Details.

The Request Time vs. Think Time Chart

The Request Time vs. Think Time pie chart reveals the ratio of the amount of time that is required for the system to process the requests in the result set against the amount of network latency and user think time. This measurement is based on the time range that you entered in the search criteria.

This chart illustrates the impact that response time has on the productivity of users in the monitored system.

User Request Details

The User Request Detail page appears only after you click a bar in the User Requests chart on the User Requests page. No PeopleSoft portal navigation exists for this page.

User requests are the set of PMUs that together make up a complete server round trip. All of the PMUs in a user request have the same top instance ID.

This page displays additional information for every user request that appears within the User Requests chart.

Duration	The duration of the top-level PMU instance for the current user request.
Agent Start Date/Time	The date and time that the top-level PMU began according to the clock on the monitored system.
Monitor Received Date/Time	The date and time that the monitoring system inserted performance information related the top-level PMU into the monitoring database.
PMU Details	This link accesses the completed PMU component where the system automatically loads all the PMUs with the same top PMU instance into this grid.

Application Server Totals

Application Server Time	The total time spent in the application server.
Serialization Time	The time spent serializing data.
Deserializtaion Time	The time spent deserializing data.
Total Cache Misses	The total number of application server cache misses for this user request. Cache misses occur when the system must retrieve metadata from the database as opposed to the application server cache. This value is the sum of metric 1 from all PMU 412 (Tuxedo Service Summary PMU) rows with the same top PMU instance. Note. Continued high cache misses or file hits indicate cache problems that should be investigated.
Total Cache File Hits	The total number of application server cache memory hits for this user request. Cache memory hits occur when the system retrieves metadata from the application server cache file. It is the sum of metric 2 from all Tuxedo Service Summary PMU 412 (Tuxedo Service Summary) rows with the same top PMU instance.
SQL Total Time	The total time that is spent processing SQL for this user request.
SQL Exec Count	The total number of SQL statements that were run in this user request. This value is the sum of metric 6 from all PMU 400 (Tuxedo Service PCode and SQL PMU) rows with the same top PMU instance.
SQL Fetch Count	The total number of SQL fetches that were run in this user request. This value is the sum of metric 5 from all PMU 400 (Tuxedo Service PCode and SQL PMU) rows with the same top PMU instance.
Global Variable Size	The amount of memory that is used to store PeopleCode global variables. This is the sum of metric 6 from all the PMU 412 (Tuxedo Service Summary) rows with the same top PMU instance. The greater the amount of data and memory that is required for global variables, the greater the overhead that is associated with a request.
Total Cache Memory Hits	The metric that is associated with the PMU 412 (Tuxedo Service Summary). Cache memory hits occur when the system retrieves metadata from the application server memory, not the cache file or the database.

Component PMUs

This section displays these PMUs that are associated with a user request:

PMU 401 (ICPanel): The ICPanel request.

Clicking the ICPanel link launches a new browser window, which contains the component statistics for that component default from midnight to the current time.
PMU 402 (Modal Level 1): The first modal level if one exists.

The first modal level is triggered by think-time PeopleCode, such as DoModal, DoCancel, and so on.
PMU 403 (Modal Level 2): The second modal level if one exists.

The second modal level contains the modal components that are called from the first modal level, such as secondary pages or lookup pages.

PMU	The PMU label that is defined in the PMU definition. Click this link to go to the Component Statistics page for the current component so that you can compare historical performance data for that component.
Name	The name of the component containing the page that is running the request.
Page	The name of the page running the request.
Action	The component action string, passed from the web server to the application server, such as Save, Search, and so on.
Duration	Duration of the PMU on that line in the grid.
Component Buffer Size	The component buffer size metric in PMU 401 (ICPanel), 402 (Modal Level 1), and 403 (Modal Level 2). The component buffer must be deserialized by the application server when it is received from the web server, and then serialized when it is sent back to the web server. The larger the buffer, the greater the impact is of serialization overhead on the monitored system.

Queries and Scripts

This section displays the metrics of these PMUs that are associated with the user request:

PMU 410 (ICScript): metrics that are related to invoking an ICScript.
PMU 411 (ICQuery): metrics that are related to invoking an ICQuery.

PMU	The type of PMU: ICScript or ICQuery.
Name	Name of the ICScript or ICQuery.
Action	The action string, passed from the web server to the application server.
Duration	The duration of the ICScript or ICQuery.

Tuxedo Service Information

This section displays the metrics of all PMU 400s (Tuxedo Service PCode and SQL) that are associated with the current user request.

Duration	Duration of PMU 400 (Tuxedo Service PCode and SQL).
Tuxedo Service Name	The name of the Tuxedo service that was run.
Server Name	The name of the server process on the application server that is running the request.
Host/Port	The name of the server and the JSL port.

User Request Durations

The User Request Durations summarizes how time was spent servicing the user request. For example, it might show that the most time was spent in SQL.

Portal Time	The total time that was spent in the portal servlet for this request, including waits for calls to other servers. The sum of the duration of the PMU 100s (Portal Request) that are associated with the request.
PIA Time	The total time that was spent in the PIA servlet. It is a subset of the portal time if the PIA request originated from Portal. The sum of the duration of the PMU 100s (Portal Request) and PMU 106s (PIA Request From Portal) that are associated with the request.
Tuxedo Service Time	The total time that was spent in the application server. It is a subset of the PIA time for a PIA request. A large difference between the duration of the PIA bar and the Tuxedo bar indicates that the request spent time on the web server or was queued by Tuxedo. The value is equal to the sum of the duration of the PMU 400 (Tuxedo Service PCode and SQL) rows.
PeopleCode Exec Time	The total time that was spent by the application server running PeopleCode, including any SQL that was triggered from within the PeopleCode. It is a subset of the Tuxedo Service Time. The value is equal to the sum of metric 1 of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the request.
PeopleCode SQL Time	This is the total time that was spent by the application server running SQLExecs and SQL objects that were triggered from within the PeopleCode. It is a subset of the PeopleCode Exec Time. The value is equal to the sum of metric 2 of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the requests.
PeopleCode Built-in SQL Time	This is the total time that was spent by the application server running SQL that was triggered from within PeopleCode built-in functions such as GetNextNumber(). It is a subset of the PeopleCode Exec Time. The value is equal to the sum of metric 3 of the PMU 400s (Tuxedo Service PCode and SQL) that are associated with the requests.
PeopleTools SQL Exec Time	This is the total time that was spent by the application server running SQL that was triggered by the PeopleTools runtime. It is a subset of the Tuxedo Service Time. It may overlap with the PeopleCode Exec time. The value is equal to the sum of metric 4 of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the requests.

Note. The number of bars appearing in the graph vary per request type.

PeopleCode Durations

This chart is a summary showing which PeopleCode event types (such as SavePreChange, SavePostChange) were triggered for this user request, how many times each type was triggered, and how much time was spent in running PeopleCode for each event type.

The PeopleCode Durations chart contains data only if a user accessed a component or ran an iScript while the filter level was set at verbose. The data is derived from PMUs 500–516.

These PMUs contain metrics for the cumulative execution count and execution time at the record field, component record, page, and component-level for each PeopleCode event type.

The number of times that a particular PMU ran appears in parenthesis.

All charts have a View in Grid hyperlink to show the chart information in a grid. This enables you to sort the information and download the information to Excel spreadsheets.

Viewing Component Trace Information

Access the Component Trace page.

To view the Component Trace page, a user must have completed a performance trace. This page enables you to isolate the component performance information within a particular performance trace. The information presented appliesonly to component processing.

Note. To gain the most performance data and a clearer insight into component activity, you should run the performance trace that is generating the trace information in debug or verbose mode.

Performance Trace Name	Displays the name of the trace, which is typically the end user's user ID, but can be whatever name was entered by the end user in the Performance Trace Name edit box on the Performance Trace Console.
Monitor Start Date and Time	Displays the date and time that the monitor began receiving performance information for a particular performance trace.
First Component	Identifies the first component that is accessed during the performance trace.
Component Cache Status	Enables you to determine whether the objects that are associated with a component were accessed from the application server cache or retrieved from the database during the performance trace.
Server Round Trips	Displays the number of server round trips that are related to component processing that were completed during a particular performance trace.
SQL Executes	Displays the number of SQL programs that were run during component processing.
SQL Fetches	Displays the number of SQL fetches that were run during component processing.
PeopleCode Program Executions	Displays the number of PeopleCode programs that were run during component processing.
Email XML File	Enables you to email the data from the Component Trace page in an XML format so that you can share the information with colleagues or PeopleSoft consultants.

Warnings

The Warnings section contains any warnings that may alert you to situations in the trace that may cause discrepancies in or misunderstandings of the trace results. For example, if particular component objects were not fully cached at the time of the trace, a warning appears notifying you of this. Production systems usually cache all frequently accessed objects for better performance. Objects that are not cached can skew results because after an object is fully cached, performance when accessing that object typically increases dramatically. The Description column provides details of the warning and provides considerations or remedies where appropriate.

A warning message appears if the performance trace is not run in debug mode.

Duration Summary

You see a duration summary if the user ran the performance trace with standard or higher filter level.

The Duration section contains information regarding the duration of the entire performance trace as well as other elements of the performance trace, such as the time spent processing SQL and the time spent processing PeopleCode. For example, the duration information may show that the time for the total trace was 50 seconds, and 22 seconds were spent processing SQL while 15 seconds were spent processing PeopleCode.

Note. The duration total is not a sum of the measurements appearing here. The PeopleSoft system does not instrument every aspect of the runtime, and these areas that are not instrumented account for the difference between the total duration and the total of the measurements that are displayed here.

Measurement	Displays the measurement appearing in the row, such as SQL, PeopleCode, or Total Trace.
Duration	Displays the total time that was consumed by each element of a performance trace. For example, it may show the total trace consuming 50 seconds and 22 seconds were attributed to SQL.
% of Total	Displays the percentage of the total trace value that each element of the trace consumed. For example, if the total trace value is 50 seconds and the SQL value is 22 seconds, the % of total value is 44.

Server Round Trips

The Server Round Trip section contains information for each server round trip that was completed during a particular performance trace. It contains two tabs: Durations and PeopleTools State Management.

You see a duration summary if the user ran the performance trace with standard or higher agent filter level.

The Durations tab contains the following information:

Seq (Sequence)	Displays the sequential order in which the server round trips were run, with 1 indicating the first round trip of the performance trace.
Action	Displays the action that was completed during the server round trip, such as Load Page/Search Page, Load Search Result, Click PeopleCode Command Button for field FUNCLIB <table.field>. This column is a hyperlink that launches the Round Trip Details page.
Component	Displays the component from which the user invoked the action.
Page	Displays the page within the component from which the user invoked the action.
Duration	Displays the amount of time that each action, or server round trip, consumed.
% of Total Trace	Displays the percentage of the total performance trace duration that each server round trip consumed. For example, if the total trace value is 50 seconds and a particular action consumes 3.5 seconds, the % of total trace is 7.

The PeopleTools State Management tab contains the following additional information:

Pack/Unpack Time

Reflects the time that was required for serialization (pack) and deserialization (unpack). The value is derived as follows:

PMU 427 + 428 duration - (time spent performing SQL in 427 and 428)

The value reflects the serialization duration plus the deserialization duration minus any time spent performing SQL as part of serialization and deserialization.

Component Buffer Size

The component buffer size, metric 1, in PMU 401 (ICPanel). The component buffer has to be deserialized by the application server when it is received from the web server, and then serialized when it is sent back to the web server.

The larger the buffer, the greater is the impact of serialization overhead on the monitored system.

PeopleCode Global Size

This is the size of the memory that is used for PeopleCode global variables, which are sent to and from the application server during a Jolt request. This is included in the send buffer size and receive buffer size.

This is metric 5 for all PMU 412 (Tuxedo Service Summary) rows.

SQL Summary

The SQL Summary section contains additional details regarding the SQL activity during component processing. It contains two tabs: SQL Executes and SQL Fetches.

You see a SQL summary if the user ran the performance trace with verbose or higher agent filter level.

The SQL Executes tab contains the following information:

Seq (Sequence)	The system orders the rows by total duration. The sequence in this case is a ranking of total duration.
SQL Operation and Tables	Displays the SQL operation and tables (such as UPDATE PS_QUERY_RUN_QRYVW).
SQL Type	Reflects the type of source code making the SQL call. The SQL types are: PeopleTools (PMU 408). Inline PeopleCode (PMU 406). PeopleCode Built-In (PMU 407).
Execution Count	Displays the number of times that the SQL was run during the performance trace.
Total Duration	The total time required by all executions.
Total Duration Average	The average execution duration, as in Total Duration / Execution Count.

The SQL Fetches tab contains the following additional information for each SQL statement that was run:

Fetch Count	The total count of all SQL fetches that were run during component processing within a particular performance trace.
Fetch Count Average	The average number of fetches per SQL Execute.
Fetch Duration	The sum total of all durations for SQL fetches.
Fetch Duration Average	The system average for SQL fetch durations during component processing.

PeopleCode Summary

The PeopleCode Summary section contains additional details regarding the PeopleCode activity during component processing. These values relate to PMU 404 (PeopleCode Program Execution).

You see a PeopleCode summary only if the user ran the performance trace in debug mode.

The following metrics apply only to the performance information that was collected during the performance trace.

Seq (Sequence)	The system orders the rows by .
PeopleCode Program	Displays the name of the PeopleCode program that was run.
Execution Count	Displays the number of times this program has been run.
Duration Total	Displays the sum of all durations of the PeopleCode program.
Duration Average	Displays the average duration of the PeopleCode program.
Duration Maximum	Displays the longest duration of this PeopleCode program.
Duration Minimum	Displays the shortest duration of this PeopleCode program.

Round Trip Details

The Round Trip Details page contains a subset of the performance data that appears on the Component Trace page. It enables you to drill down into a specific round trip to gather duration, SQL, and PeopleCode information. The data that is presented in the page is similar to the data in the Performance Trace page. The PMU Details link accesses the completed PMU page with all PMUs that were captured in this action.

Viewing Component Statistics

Access the Component Statistics page.

The Component Statistics page enables you to gather performance information that is related to a specific PeopleSoft component.

System ID	Identifies each monitored system. PeopleSoft automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database that is running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The user ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the Lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Component	Select the component name.
Market	Select the market to which the component belongs.
Search	The Search button performs a search based on the search criteria that is defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Refresh	Updates the To Time value to equal the current time. This enables you to view the most current results of a search.
Component Execution Count	The number of times that a component was accessed within the specified time period. This value is calculated by selecting the count of PMU 400 (Tuxedo Service PCode and SQL) rows that were found with the current search results.

Component Averages

The Component Averages section displays the average and standard deviation of the following metrics.

Duration

The average duration of the PMU 400 (Tuxedo Service PCode and SQL) rows that were run from this component.

SQL Fetch Count

The average number of SQL fetches that were run for this component.

The value is the average of metric 5 from all PMU 400 (Tuxedo Service PCode and SQL) rows with Context 1 equal to this component's name.

SQL Execute Count

The average number of SQL statements that were run for this component. The average of metric 6 from all PMU 400 (Tuxedo Service PCode and SQL) rows with Context 1 equal to this component's name.

Component Sizes

The Component Sizes chart displays the average and standard deviation of the component buffer sizes.

Receive Buffer Size	The average size of the data that the application server received from the web server during a JOLT request. This is metric 2 for PMU 115 (JOLT Request).
Send Buffer Size	Size of the data that was sent from the web server to the application server during a JOLT request. This is metric 1 for PMU 115 (JOLT Request).
Component Buffer Size	This is the average component buffer size (metric 1) in PMU 401 (ICPanel). The component buffer has to be deserialized by the application server when it is received from the web server, and then serialized when it is sent back to the web server.
PeopleCode Global Size	This is the average size of the memory that is used for PeopleCode global variables, which are sent to and from the application server during a JOLT request. This is included in the send buffer size and receive buffer size. This is metric 6 for all PMU 412 (Tuxedo Service Summary) rows.

Component Durations

This chart shows the average and the standard deviation of these metrics:

Portal Time	The average time that was spent in the portal servlet for this component, including waits for calls to other servers. The value is the duration of the PMU 100 (Portal Request) rows that are associated with requests that are accessing this component.
PIA Time	The average time that was spent in the PIA servlet for this component. The value is the duration of the PMU 100 (Portal Request) rows and PMU 106 (PIA Request From Portal) rows that are associated with the request accessing this component.
Tuxedo Service Time	The average time that was spent in the application server for this component. The value is the duration of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the request that is accessing this component.
PeopleCode Exec Time	The average time that was spent by the application server running PeopleCode, including any SQL that was triggered from within the PeopleCode. The value is the average of metric 1 of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the component.
PeopleCode SQL Time	This is the average time that was spent by the application server running SQLExecs and SQL objects that were triggered from within the PeopleCode. It is a subset of the PeopleCode Exec Time. The value is equal to the average of metric 2 of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the requests. Note. This is the average time that was spent for running all SQL statements of this type in a single application server request, not the average duration of one SQL statement of this type.
PeopleCode Built-in SQL Time	This is the average time that was spent by the application server running SQL that was triggered from within PeopleCode built-in functions such as GetNextNumber(). It is a subset of the PeopleCode Exec Time. The value is the average of metric 3 of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the requests. Note. This is the average time that was spent for running all SQL statements of this type in a single application server request, not the average duration of one SQL statement of this type.
PeopleTools SQL Exec Time	This is the average time that was spent by the application server running SQL that was triggered by the PeopleTools runtime. It is a subset of the Tuxedo Service Time. This value may overlap with the PeopleCode Exec time. The value is the average of metric 4 of the PMU 400 (Tuxedo Service PCode and SQL) rows that are associated with the requests. Note. This is the average time that was spent for running all SQL statements of this type in a single application server request, not the average duration of one SQL statement of this type.

PeopleCode Durations

The PeopleCode Durations chart contains data only if a user accessed a component while the monitoring level was set at verbose. The data is derived from PMUs 500–516.

This chart is a summary showing:

Which PeopleCode event types (such as SavePreChange and SavePostChange) were triggered for this component at the record field, component record, page, and component levels.
The average time that was spent running each event instance.
The sample count that was used to calculate these averages.

The standard deviation is not calculated for this chart.

Viewing Portal Statistics

Access the Portal Statistics page. This page enables you to view the performance of the portal servlet.

All charts have a View in Grid hyperlink to show the chart information in a grid. This enables you to sort the information and to download to Microsoft Excel spreadsheets.

The chart on this page is generated from all Portal Request PMUs 100 (Portal Request) rows grouped by action (login, logout, homepage, target page, error), all PMU 117 (GetContent) rows, and all PMU 116 (Redirect after Login) rows.

Before you view the charts, you must enter search criteria.

System ID	Identifies each monitored system. PeopleSoft automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that was generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the Lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Search	The Search button performs a search based on the search criteria that is defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Refresh	Updates the To Time value to equal the current time. This enables you to view the most current results of a search.

Portal Action Durations

This chart reveals the average and standard deviation of the durations of the following metrics. The sample size that is used to derive the average appears in parentheses to the right of the bar label.

Redirect	The time that was required for the browser to load a page that is requested by way of a URL redirection.
Get Content	The time that was required for the system to display target content.
Portal View page	The time that was spent to assemble a portal page, excluding the homepage.
Portal login	The time that was spent on processing a login request.
Portal Homepage	The time that was spent assembling the homepage.
Portal error	The time that was spent assembling the error reporting page.
Portal Expire	The time that was spent processing a portal page expiration.
Portal Logout	The time that was spent on processing a logout request.
Portal Refresh Pagelet	The time that was spent refreshing a pagelet. A pagelet refresh gets triggered by the refresh button on a pagelet from the homepage.

Viewing PIA Statistics

Access the PIA Statistics page.

This page enables you to view the performance of the PIA servlet.

All charts have a View in Grid hyperlink to show the chart information in a grid. This also enables users to sort the information and to download to Microsoft Excel spreadsheets.

System ID	Identifies each monitored system. The PeopleSoft system automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the Lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Search	The Search button performs a search based on the search criteria that is defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Refresh	Updates the To Time value to equal the current time. This enables you to view the most current results of a search.

PIA Action Durations

The average time for a PIA servlet to handle a direct request from a browser (PIA View Page), and handle a request from a Portal.

The data is generated from PMU 101 (PIA Request), and the data is grouped by context 3, which is action (login, logout, homepage, target page, error). Additional performance data is generated from PMU 102 (Compression Request), PMU 106 (PIA Request From Portal), and PMU 114 (Attach File Request).

PIA Action Response Sizes

The average size of the content sent to a browser when fulfilling a PIA request for the same rows of data that appear in the PIA Action Durations chart.

The data is generated from the metric 1 values for PMU 101 (PIA Request), metric 1 values for 106 (PIA Request from Portal), metric 1 values for PMU 114 (Attach File Request), and metric 1 values for PMU 102 (Compression Request).

Viewing Top Portal Content Requests

Access the Top Portal Content Requests page.

This page enables you to identify, view, and rank the performance data that is associated with individual requests for content through the PeopleSoft portal. The charts enable you to identify content requests that are both popular and performing poorly.

All charts have a View in Grid hyperlink to show the chart information in a grid. This enables you to sort the information and to download to Microsoft Excel spreadsheets.

Before you can view portal content requests, you must first enter search criteria.

System ID	Identifies each monitored system. PeopleSoft automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	Enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the Lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Search	The Search button performs a search based on the search criteria that is defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Number to Retrieve	The number of top ranked statistics to be displayed on the charts.
Refresh	Updates the To Time value to equal the current time. Enables you to view the most current results of a search.

Viewing Top PeopleCode Events

Access the Top PeopleCode Events page.

These charts show the PeopleCode events that have highest durations and execution counts, grouped by component and event type. Durations and counts are summed across levels (record field, component record, page and component) for each component and event type.

All charts have a View in Grid hyperlink to show the chart information in a grid. This also enables users to sort the information and to download to Microsoft Excel spreadsheets.

Note. Performance data must be captured in verbose agent filter mode or higher to be displayed.

Before you can view PeopleCode events, you must first enter search criteria.

System ID	Identifies each monitored system. PeopleSoft automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the Lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
Component	Select the component name.
Market	Select the market to which the component belongs.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Search	The Search button performs a search based on the search criteria that is defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Number to Retrieve	The number of top ranked statistics to be displayed on the charts.
Refresh	Updates the To Time value to equal the current time. Enables you to view the most current results of a search.

Note. If the filter level of the PSAPPSRV agent type is set to a level that is lower than the level of PMU 500–516 (Verbose), a warning message appears at the top of the page indicating that information for the current component may not be available. This data is available only in the following situations: the agent filters on the application server are set to verbose or debug, or performance traces were run with a filter level set to debug or verbose. A small, yellow triangle appears at the top of the page alerting you if application servers for that system are currently at a higher filter level than verbose. This, however, does not mean that you would not see data; PMUs may exist from performance traces or lower filter levels from past monitoring.

Top PeopleCode Events

This chart indicates the PeopleCode events that are run most often.

It shows the highest counts for all PMU 500–516. The counts consist of adding metrics 1, 3, and 5. The data is grouped by component and event name (contexts 1 and 2).

See Setting Agent Filter Options.

Top Cumulative Durations

This chart reveals the largest sum of all the durations of a specific PeopleCode event.

It shows the top sum of PMU 500–516 durations. The duration is not the duration of the PMU; the duration is the sum of metrics 2, 4, and 6. The data is grouped by component and event name (contexts 1 and 2).

Duration Averages

This chart shows the components with event types that on average consume the most time in application server requests.

Duration averages are the averages and standard deviations of the sum of durations of PMU 500–516. The duration is not the duration of the PMU; the duration is the sum of metrics 2, 4, and 6. The data is grouped by component and event name (contexts 1 and 2). The sample count that is used to calculate these averages is not the execution counts of the events (metrics 1, 3 and 5) but the number of times this component reported that event type.

Viewing Top PeopleCode Executions

Access the Top PeopleCode Executions page.

This page enables you to identify the PeopleCode programs that the system runs the most and the performance of the PeopleCode programs.

The data on this page is generated from PMU 404 (PeopleCode Program Execution), and it is based on all PeopleCode function calls and all calls to application packages.

All charts have a View in Grid hyperlink to show the chart information in a grid. This enables you to sort the information and to download to Microsoft Excel spreadsheets.

Note. Performance data must be captured in debug agent filter mode to be displayed in this chart.

Before you can view top PeopleCode executions, you must first enter search criteria.

System ID	Identifies each monitored system. PeopleSoft automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
Component	Select the component name.
Market	Select the market to which the component belongs.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Search	The Search button performs a search based on the search criteria that is defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Number to Retrieve	The number of top ranked statistics to be displayed on the charts.
Refresh	Updates the To Time value to equal the current time. Enables you to view the most current results of a search.

Note. This data is available only when the agent filters on the application server are set to debug, or performance traces were run with a filter level set to debug. A small, yellow triangle appears at the top of the page alerting you if application servers for that system are currently at a higher filter level than debug. This, however, does not mean that you would not see data; PMUs may exist from performance traces or lower filter levels from past monitoring.

Top PeopleCode Executions

Indicates the PeopleCode programs that the system runs the most.

Top PeopleCode executions refers to the highest counts for all PMU 404 (PeopleCode Program Execution) rows. The data is grouped by component and program name (contexts 1 and 2).

Top Cumulative Durations

Identifies the PeopleCode programs with the longest cumulative durations. The duration of each execution of a particular PeopleCode program is recorded and added to the sum total to create a cumulative duration value.

The data is generated from the sum of PMU 404 (PeopleCode Program Execution) durations, and the data is grouped by component and program name (contexts 1 and 2).

Duration Averages

Identifies the PeopleCode programs with the highest average duration in the system.

The data is generated from the sum of PMU 404 (PeopleCode Program Execution) durations, and the data is grouped by component and program name (contexts 1 and 2).

Viewing Top PeopleCode SQL Statements

Access the Top SQL Statements page.

The data for this page is generated from the sum of all PMU 406 (PeopleCode SQL Execute) rows, and the data relates to the SQL that PeopleCode programs run.

The bars on the chart are links that you can click to access the Completed PMU history component. The system automatically populates the search criteria with values derived from the Top PeopleCode SQL Statements page. For example, PMU Identifier is set to 406, and the Context 2 and Statement Number values are set appropriately.

All charts have a View in Grid hyperlink to show the chart information in a grid. This also enables you to sort the information and to download to Microsoft Excel spreadsheets.

System ID	Identifies each monitored system. PeopleSoft automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the Lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
Component	Select the component name.
Market	Select the market to which the component belongs.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Search	The Search button performs a search based on the search criteria that is defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Number to Retrieve	The number of top ranked statistics to be displayed on the charts.
Refresh	Updates the To Time value to equal the current time. Enables you to view the most current results of a search.

Top SQL Statements

This chart indicates the PeopleCode SQL statements that the system runs the most.

SQL statements are identified by PeopleCode locations, such as PeopleCode Program and PeopleCode Statement Number (context 2 and metric 2).

Top Cumulative Durations

Identifies the PeopleCode SQL statements with the longest cumulative durations.

SQL statements are identified by PeopleCode locations, such as PeopleCode Program and PeopleCode Statement Number (context 2 and metric 2).

Duration Averages

Identifies the PeopleCode SQL statements with the longest average durations.

SQL statements are identified by PeopleCode locations, such as PeopleCode Program and PeopleCode Statement Number (context 2 and metric 2).

Note. This data is available only when the agent filters on the application server are set to verbose or debug, or performance traces were run with a filter level set to debug or verbose. A small, yellow triangle appears at the top of the page alerting you if application servers for that system are currently at a higher filter level than verbose. This, however, does not mean that you would not see data; PMUs may exist from performance traces or lower filter levels from past monitoring.

Viewing Top Components

Access the Top Components page.

All charts have a View in Grid hyperlink to show the chart information in a grid. This enables you to sort the information and to download to Microsoft Excel spreadsheets.

The data on this page is generated from PMU 401 (ICPanel), and it is grouped by action, as in ICSave. The system passes the "action" to the application server as part of an ICPanel service request specifying the reason for the request.

On this page, the charts are grouped by action. The "action" refers to the action of a user in a component. The following are examples of how the actions appear:

"#KEY" translates into Load Component.
"#ICPanel1" translates into "Tab to Page."
Launch Page/Search Page.
Click PeopleCode Command Button for Field DERIVED_PSPM.PM_FILTERCHANGE.
Switch to Add Action Mode.
Launch Secondary Page.
Change Drop Down List Value for Field PRCSRQSTDLG_WRK.RECURNAME.
Click OK or Save to Save Secondary Page.
Click Cancel on Secondary Page.
Save Component.
Select Row0 From Search Result.
Select Row1 From Search Result.
Sort Column on Grid.
Tab to Next Page.
View Page.

Before viewing top components, you must first enter search criteria.

System ID	Identifies each monitored system. PeopleSoft automatically generates this value incrementally. System definitions are created automatically when the first agent of a monitored system registers with the monitoring system.
Database Name	The name of the PeopleSoft application database running on the monitored system. The monitoring system automatically inserts this value when it recognizes and creates a monitored system.
User ID	The user ID of the user who generated the request. The User ID field prompts against the PSPMOPRDEFN table, which is populated by the Lookup program (PSPM_LOOKUP).
Performance Trace Name	You enter the performance trace name to search on PMUs within a particular performance trace. If a user has launched a performance trace in the Performance Console, the system labels every PMU that is generated in a business process with a performance trace name. Performance Trace Name prompts against the PSPMPERFTRACE table, which is populated by the Lookup program (PSPM_LOOKUP). Note. When searching using a performance trace name, the search ignores all agent IDs, so the current system ID is not relevant.
Component	Select the component name.
Market	Select the market to which the component belongs.
From/To Date/Time	Specify a range of time. The system searches on the monitor date and time, not the agent date and time. If you do not specify a time range, by default, the system displays information for the current day, starting at midnight (12:00 AM).
Search	The Search button performs a search based on the search criteria defined on the page. It should be pressed only once per search. If you want to view the most current information, use the Refresh button after running the initial search. If you click Search repeatedly, the system returns the same results from the same time range for each search.
Number to Retrieve	The number of top ranked statistics to be displayed on the charts.
Refresh	Updates the To Time value to equal the current time. This enables you to view the most current results of a search.

Top Components

This chart indicates the components that users access and use the most.

The data is grouped by component, page, and action (contexts 1, 2, and 3).

Top Cumulative Durations

Identifies the component PMUs with the longest cumulative durations.

The data is grouped by component, page, and action (contexts 1, 2, and 3).

Duration Averages

This chart displays the average duration and standard deviation of the PMUs that are associated with a particular component.

The data is grouped by component, page, and action (contexts 1, 2, and 3).