-
Cisco Application Velocity System User Guide (Software Version 5.0)
-
Preface
-
Product Overview
-
AVS Description
-
Appliance Administration
-
Command Line Interface
-
Configuration Reference
-
AppScreen Configuration
-
Management Console
-
Reporting
-
NMS Integration
-
Availability Manager Clustering
-
Database Maintenance
-
Logs
-
SNMP MIB
-
Deployment Options
-
Frequently Asked Questions and Troubleshooting
-
Anonymous Base File Statistical Model
-
Regular Expressions
-
AppScreen Rules DTD
-
Glossary
-
Index
-
Table Of Contents
Accessing the Management Console
Performance Monitoring Details
Drilldown by Source IP Address Blocks
Drilldown by Defined Source Locations
Configuring Email for Scheduling
Managing Transaction Type Mapping
Defining Business Transactions
Editing a Business Transaction
Deleting a Business Transaction
Reporting
Note
If you have installed only a Cisco AVS 3120 Application Velocity System, reporting functions are not available and you will not see a Reports folder in the menu at the left side of the Management Console window. You must be running the Management Console on a Cisco AVS 3180 Management Station in order to see the Report items in the Management Console. This note does not apply to users who have performed software upgrades to AVS 5.0 from older software products or on the Velocity Appliance.
This chapter discusses how to use the Management Console reporting features. For general information about using the Management Console for management and administration, see "Management Console." This chapter discusses the following topics:
•
•
For details on the Database Archiving command, see "Database Maintenance."
Note
Accessing the Management Console
To view the Management Console, you'll need to know its IP address and port number. Start your web browser and enter a URL in the following format:
http://consoleIPAddress:consolePort/fgconsole/For example:
http://10.0.0.2:9000/fgconsole/A dialog will request your user name and password. The default values are:
user name: admin
password: adminFor details on changing the user name and/or password, refer to "Accessing the Management Console" section.
After entering the correct user name and password, the web browser displays the Management Console shown in Figure 8-1.
Figure 8-1 Management Console Main Page
For details on creating clusters, adding nodes to a cluster, and other items not covered here, see "Management Console."
Note that before you can manage an individual application appliance, you must create at least one cluster and add the application appliance node to it. After you have added one or more nodes to a cluster, their names will appear at the bottom of the list of cluster menu items, and you can click a node name to manage the individual node.
The following sections describe the specific functions related to reporting and management that you can perform by using the menu options in the left pane under the Reports folder.
Note
Bandwidth Savings Reports
The Bandwidth Savings Reports page (shown in Figure 8-2) allows you to generate color-coded graphical reports, based on a configurable reporting period, that show the actual bandwidth savings realized for traffic delivered by the application appliance. In addition, this feature provides both summary reports and detailed reports organized by content type to enable you to easily quantify actual bandwidth reduction and the resulting return on investment (ROI).
Figure 8-2 Bandwidth Savings Reports Page
The Bandwidth Savings Reports page contains controls to set the report time period and domain filter. The time period can be set either as recent activity in the last minutes/hours/days or as a date range.
You can enter a specific domain (such as "www.yahoo.com" or "usatoday") to report bandwidth savings just for that domain. By default, the Domain Search String is set to "All Domains," which generates a report for all domains.
Additionally, you can choose your output format: HTML or PDF.
An example Bandwidth Savings Report is shown in Figure 8-3. All domains are selected. Graphs compare bandwidth for uncondensed and condensed traffic, and show bandwidth usage for various content types.
Figure 8-3 Bandwidth Savings Report
Throughput Reports
The Throughput Reports page (shown in Figure 8-4) allows you to generate graphical reports, based on a configurable reporting period and reporting interval. The reports show application appliance throughput performance as measured in both transactions per second and KB per second.
Figure 8-4 Throughput Reports Page
The Throughput Reports page contains controls to set the report time period and report sampling interval. The time period can be set either as recent activity in the last minutes/hours/days or as a date range.
The report (sampling) interval can be set to a number of minutes or hours. This specifies the smaller periodic interval during which the average throughput is calculated. For example, if you specify a one day report range, and a report interval of two hours, data for each two-hour block during the day is averaged together, and 12 (24 divided by 2) data points are graphed along the X axis.
Additionally, you can choose your output format: HTML or PDF.
An example Throughput Report is shown in Figure 8-5. The report displays throughput in two different graphs as transactions per second and KB per second.
Figure 8-5 Throughput Report
AppScreen Reports
The AppScreen Reports folder within the Reports menu includes commands that allow you to generate reports of requests that matched AppScreen Classes. You can use these reports to monitor attempts at web site intrusion and application security breach. For more details on AppScreen request filtering, see "AppScreen Configuration."
Click on the AppScreen Reports menu folder to see the following reporting commands:
These commands are described in the following subsections.
Graphical Incidents Summary
The Graphical Incidents Summary page (shown in Figure 8-6) allows you to generate a graphical pie chart, based on a configurable reporting period. The pie chart represents all the requests during the reporting period that matched AppScreen Classes. The proportion of requests of each severity level is shown by the colored pie slices in the chart.
To generate a chart, set the Start and End dates and times and click Update. Note that the Start and End dates and times are initially filled in with default values covering the last month, up to the present time, so you don't need to enter anything if this period is acceptable.
Figure 8-6 AppScreen Reports Graphical Incidents Summary Page
You can click on any one of the colored pie slices to see a report of the incidents at that particular severity level. For more information on this report, see the following section, Tabular Incidents Details.
Tabular Incidents Details
The Tabular Incidents Details page (shown in Figure 8-7) allows you to generate a tabular report of AppScreen incidents, based on a configurable reporting period. The table includes all the requests during the reporting period that matched AppScreen Classes. The table is sorted by severity, with the most severe incidents (level 1) at the top.
To generate a report, set the From and To dates and times and click Submit. Note that the From and To dates and times are initially filled in with default values covering the last month, up to the present time, so you don't need to enter anything if this period is acceptable.
You can limit the incidents included in the report by setting the following other controls on this page:
•
•
•
Figure 8-7 AppScreen Reports Tabular Incidents Details Page
You can click the Hide Criteria link at the top of the page to hide the controls by which you set the report criteria. If the controls are hidden, click Show Criteria to show them.
Incident Details
The Incident Details page (shown in Figure 8-8) allows you to display details on a particular AppScreen incident, based on the incident ID.
To retrieve the details on an incident, enter the incident ID in the Incident ID field and click Submit. Or, you can click on an incident URL in the tabular summary of incidents report to display details on that incident.
The incident ID can be obtained from the Incident ID field in the SNMP trap, if you configured AppScreen to send traps to an SNMP Management Station.
Figure 8-8 AppScreen Reports Incident Details Page
An example of an incident details report is shown in Figure 8-9.
Figure 8-9 AppScreen Reports Incident Details Example
AppScope Reports
The AppScope Reports page (shown in Figure 8-10) enables you to generate reports that measure true end-to-end application performance as seen by real end users. AppScope also accurately determines both the server delay and network delay components associated with the user experience at the transaction level. AppScope's unique Statistical Traffic Sampling technology enables an enterprise to statistically sample user requests, making AppScope highly scalable for high-traffic applications.
You can measure true application performance by grouping transactions into sets of business transactions that define a series of individual web page transactions done by a user in a typical application task such as filing an online expense report. Aggregate statistics and charts show performance trends for the whole business transaction in addition to individual transactions. For more details on business transaction, see the "Defining Business Transactions" section.
Performance Monitoring Details
The AppScope performance monitoring feature measures total page download times from the point of view of the client. AppScope measures both accelerated (condensed) and pass-through (uncondensed) transactions, so that one can compare these and determine the benefit of application appliance optimization. When measuring pass-through transactions, AppScope operates in proxy mode.
AppScope's ability to measure a sample of all transactions enables it to easily scale in high-traffic situations. When used in a load-balanced environment, the load-balancer sends some portion of the traffic to AppScope (via the load balancing rule). In this scenario, AppScope should be configured to measure 100% of all transactions it sees. In other words, although AppScope will see only a portion of the traffic, it should be configured to measure all of this portion.
In the case where AppScope is not used with a load-balancer (that is, all transactions are passed through the AppScope server), AppScope should be configured to measure only a portion of the traffic. Although AppScope can be configured to measure 100% of the traffic in this scenario, for scalability reasons we recommend measuring only a portion of the traffic. For example, it can be configured to measure 10% of accelerated transactions and 10% of pass-through transactions, as follows:
AppScopeOptimizeRatePercent 10
AppScopePassThruRatePercent 10When measuring a transaction, AppScope begins measuring the time when the server first receives the main request for a page. It stops measuring time when the client acknowledges that it has received the remaining bytes of data for the page, including all embedded objects.
In order to accurately measure true page download time from the client's perspective, AppScope inserts a small amount of JavaScript code into the page that is delivered to the client. This code calculates the page download time on the client and reports it back to the server via a query parameter in a GET request. The server inserts the performance data into a log file, which is then loaded into the Management Console database, and from which performance reports are generated.
Measurement sampling can be configured in two ways:
•
Each transaction request is treated individually and randomly selected for a measurement group: accelerated measurement, pass-through measurement, or no measurement.
•
Once a transaction request has been randomly selected for a measurement group, all subsequent requests from that client are handled the same, for a configurable period of time, up to one day.
This feature is configured by the AppScopeCookieTTL directive, discussed in Table 5-1. When set to 0, AppScope uses individual request sampling; when set to a number of seconds, AppScope samples in sessions of that length. This is done through the FGNPERFMON cookie, which informs AppScope of the client's assigned measurement group, until it expires and is deleted.
AppScope performance monitoring relies on very accurate time measurement, in the millisecond range. If you have installed multiple application appliances, different parts of a single transaction could be handled by different nodes, and thus it's important that the clocks be synchronized. To satisfy this requirement, you should enable NTP (Network Time Protocol) on all application appliance nodes.
Basic Reports
This section describes how to generate basic AppScope Performance reports.
Tip
Click on the AppScope Reports item in the left menu to display the main AppScope reports page, as shown in Figure 8-10.
Note
Figure 8-10 AppScope Main Page
The buttons across the top of the window provide access to the main functions:
•
•
•
Tip
If you are not at the AppScope main page, click the Query button to go there.
Defining the Query
The query page contains a number of controls that set the query parameters for generating a report:
•
•
–
–
–
•
•
•
•
•
•
•
•
Date Selection
In the Start Date and Time controls, you choose a starting time for the report. You can choose a relative time in the past (relative to when the report is run), or a specific date. The available choices are shown in Table 8-1.
If you choose one of the first five relative times, the date input field changes as appropriate to let you specify the value for "N." For example, if you choose "N minutes ago," the date input field allows you to enter a number of minutes.
Generating the Report
Click the Submit button to submit the query to the database and show the report. You can click the Reset button to reset all Query form fields to their default values.
Figure 8-11 shows an example of a report resulting from a query showing pass-through requests and organizing the data by URLs.
Figure 8-11 Pass-Through Requests by URL Report
The URLs requested are shown down the left column in alphabetical order. Next to each URL, the columns to the right show the following performance data:
•
•
•
•
•
•
The same kind of data is shown for a report on accelerated page requests.
You can click on any column to sort all the report rows by the data in that column. You can toggle the sort order between ascending and descending by clicking again on the same column. The red arrow indicates the sort column and sort order (up for ascending and down for descending).
You can save any report by clicking the small disk icon
at the top of the report page. You might want to save a report so that you can regenerate it at a later time. For more details about saved reports, see the "Saved Reports" section.
To chart the data for any row and see data trends, click on the chart icon
shown at the left end of each row. For more details, see the "Charts" section.
You can request each report or chart in Adobe Acrobat (PDF) format. To do so, click the Adobe Acrobat icon
that appears on the report or chart page.
Figure 8-12 shows another example of a report where Organize Output By is set to Business Transaction. In this kind of report, each row represents a business transaction rather than a URL. A row shows the average aggregated performance data for the business transaction.
Figure 8-12 Business Transaction Report
Finally, you can generate a report that compares accelerated and pass-through performance data, as shown in Figure 8-13.
Figure 8-13 Comparison Report
The URLs (or Application classes) are shown down the left column in alphabetical order. Next to each URL, the columns to the right show the following performance data:
•
•
•
•
•
•
Note
Drilldown Reports
After generating a basic report, you can generate drilldown reports from it.
On the basic report, you can click on a URL, Application class, or business transaction in the left column to generate a drilldown report on that item. Depending on what you selected on the Query page, you can drill down by Source IP Address Blocks or by Defined Source Locations.
For drilling down into business transactions, when you click on a business transaction name, it shows a new report (see Figure 8-14) listing the individual transaction types in the business transaction, then clicking on one of those drills down into the individual URL, and finally into the source IP block or defined location as usual.
Figure 8-14 Drilldown Report into Business Transaction
Drilldown by Source IP Address Blocks
If you selected Source IP Address Blocks for the Enable Drilldown By item on the Query page, then you can drill down multiple levels on a particular URL, Application class, or business transaction. Each drilldown level includes data from a narrower block of client IP addresses for the URL or Application class selected. For business transactions, there are intermediate levels of individual transactions and URL reports shown before IP blocks.
From a basic report organized by URL, clicking on a URL drills down one level, showing a report similar to that in Figure 8-15, where IP address blocks are shown on each row. You can click on an IP address block up to three more times to drilldown further, narrowing the scope of the report to smaller blocks of IP addresses.
Figure 8-15 Drilldown Report by IP Address Block
Drilldown by Defined Source Locations
If you selected Defined Source Locations for the Enable Drilldown By item on the Query page, then you can drill down one level to a list of defined locations for a particular URL, Application class, or business transaction. A defined location is simply a name that has been previously defined to correspond to a range of source IP addresses. You use the Manage Locations command under the Reports folder in the menu to define a location; for details, see the "Managing Locations" section.
From a basic report organized by URL, clicking on a URL drills down one level, showing a report similar to that in Figure 8-16, where defined locations are shown on each row. The performance data is from the range of source IP addresses that corresponds to the location name. No further drilldown applies to this kind of report.
Figure 8-16 Drilldown Report by Defined Location
Charts
Charts are available for each row of data at every report level. You can use charts to view performance trends over time. Each report row includes a chart icon at the left edge of the row:
Click the chart icon to generate a chart of the row data. An example of a chart for a row of data in a business transaction report is shown in Figure 8-17. The colored stacked bars in the top chart show the various response time measurements for each reporting period. The colored bars are stacked on top of each other, with the light blue at the back, then the dark blue, light green, and dark green in the front. The red bars in the chart at the bottom show the number of hits that were measured.
Note
Figure 8-17 Chart of Transaction Data
An example of a chart of a comparison report is shown in Figure 8-18. This chart shows two colored bars for each reporting period: one for accelerated responses (green) and one for pass-through responses (blue).
Figure 8-18 Chart of Comparison Data
Filtering AppScope Data
All AppScope reports are filtered to remove outlier transactions. Outlier transactions are those that lie outside the normal range of results because they take longer than 120 seconds to complete. Outlier transactions are filtered out because they skew the averages. Typically, these anomalous transactions result from network outages, backend application issues, database issues, etc.
You may want to adjust the filter threshold to a higher or lower number than the default 120 seconds. If the application you are measuring really is slow, and you want to include such transactions in AppScope reports, you might want to set the threshold to several minutes. Or, if too many anomalies are still skewing the averages too much, you might want to reduce the outlier threshold to a lower number such as 60 seconds or less.
To change the outlier threshold value, edit this file:
$AVS_HOME/console/jboss-3.0.1_tomcat-4.0.4/server/default/deploy/fgconsole.war/properties/postgres_queries.propertiesLook for this text:
(page_time < /*<maxtime>*/ 120.0 /*</maxtime>*/)Change the value 120.0 to the number of seconds to set as the outlier threshold.
This text occurs multiple times in the properties file. Make sure to change all occurrences to the new value.
Saved Reports
This section includes the following topics related to AppScope reports:
Saving Reports
You can save any AppScope report by clicking the Save (disk) icon at the top of the report page. You might want to save a report so that you can regenerate it at a later time. This is especially useful if the report is based on a relative starting date, for example the last day, last month, or one week ago. Whenever you regenerate such a report, the report data is recalculated based on the specified time period relative to the time the report is regenerated.
When you save a report on which you have executed drilldown, sorting, or charting options, those effects are saved also, so when it is regenerated again, the report (or chart) will look exactly like it did when you saved it.
Saving reports is available only from HTML report pages, not PDF pages. It is also not available when viewing a previously saved report from the Saved Reports page or when executing a previously saved report from an external link, such as an email.
When you click the Save icon, a dialog prompts you to name the report. Enter a name and click OK, or Cancel to cancel the save. If you enter a name that is already used for a saved report, the warning shown in Figure 8-19 is displayed. Click Yes to overwrite the existing report with the new report. Click No to cancel the operation and return to the report page; in that case, click the Save icon again to choose a different name.
When you successfully save a report, the report page is redisplayed with a confirmation message at the top: "Report "Your report name" has been saved."
Figure 8-19 Duplicate Saved Report Warning
Accessing Saved Reports
To access previously saved reports, click Saved Reports under the Reports folder in the left pane of the Management Console main page. If there are saved reports, they are listed in a table, as shown in Figure 8-20.
Figure 8-20 Saved Reports List
Each report is listed on one row, followed by links that invoke various operations available for the report. The following sections describe the operations you can do on each report:
Only authenticated (logged in) Management Console users can save and manage reports using the functions described in this section. However, any user with access to the server and port that is running the Management Console can view reports as long as they have a URL to view the saved report. This allows unauthenticated users to view reports that are sent to them as links in email or saved as browser bookmarks.
Viewing Reports
You can view a saved report in HTML or PDF format by clicking either the HTML or PDF link, respectively. The report is regenerated when you click the link, so if it uses a relative starting date, the report data will be recalculated, relative to the moment that you clicked the link.
Editing Reports
You can edit the query for a saved report by clicking the Edit link. The report query page is displayed, with the selection criteria set to the original criteria used for the report. You can do three things:
•
•
•
If you modify and then save an existing report, the name field of the Save dialog is initialized with the existing name of the report. You can click OK to save the report under that name, overwriting the existing report, or you can change the name to a new name and click OK to create a new saved report. If you choose to overwrite the existing report with the same name, no additional warning is displayed.
Copying Reports
You can copy a report by clicking the Copy link. Just like for saving a report, a Save dialog is displayed where you can enter a name for the copy of the report. Click OK to save the copy, or Cancel to cancel it. You are warned if you enter a report name that already exists.
Note that another way of making a copy of a report is to edit it, view the report, click the Save icon, and specify a new name for the copy.
Scheduling Reports
The scheduling capability allows an administrator to define a periodic schedule at which a saved report is automatically generated and then emailed to a distribution list.
You can schedule a report for regeneration at a future time by clicking the Sched link.
Note
If there are one or more schedules already defined for the report, they are listed in the Report Schedule page, as shown in Figure 8-21.
Figure 8-21 Report Schedule Page
The saved report name is shown above the table. Each row in the table lists one schedule for the report. The schedule interval is shown, followed by the email distribution list, and links that allow you to edit or delete the schedule. Use the Add Schedule link below the table to add a new schedule for the report.
For more information on adding, editing, and deleting scheduling reports, see the "Scheduled Reports" section.
Renaming Reports
You can rename a report by clicking the Rename link. Just like for saving a report, a Save dialog is displayed where you can enter a new name for the report. Click OK to change the name, or Cancel to cancel it. You are warned if you enter a report name that already exists.
Deleting Reports
You can delete a saved report by clicking the Delete link. You are prompted with a dialog to confirm that you want to delete the report. Click OK to delete the report, or Cancel to leave it alone.
Scheduled Reports
The AppScope scheduled reports capability allows an administrator to define a periodic schedule at which a saved report is automatically generated and then emailed to a distribution list.
Use the Scheduled Reports command (under the Reports folder in the left pane) to view a list of all saved reports, as shown in Figure 8-22. Note that a report must be saved before it can be scheduled.
Figure 8-22 All Scheduled Reports Page
Each row in the table lists one saved report along with its schedules (if no schedule is shown, then the report is not scheduled). The report name is shown, followed by the schedule interval, and the email distribution list. The Email Configuration link above the table allows you to change the email setup.
To add another schedule to a report, edit the schedule, or delete the schedule for a report, click on the report name to display the schedules for that report, as shown in Figure 8-23.
Figure 8-23 Report Schedules Page
Each row shows one schedule for this report. Click the Edit or Delete links next to a report schedule to edit or delete the schedule. Click the Add Schedule link below the table to add a schedule for the report.
•
Adding a Schedule
You can add a schedule for a report by clicking the Add Schedule link below the schedule table. The schedule configuration page is displayed.
Initially, the Report Frequency is set to Weekly, which means the schedule is for a weekly report. You can change this to one of the other choices for a repeating schedule: Daily, Monthly, or Yearly. The Schedule Configurator works differently in each case and is described in the following subsections:
Each version of the configurator allows you to choose the report type. This selection is common to all versions, and thus is described here. These are the report type choices:
•
•
•
Daily Schedule
A Daily schedule is for generating the report every day. The daily configurator page is shown in Figure 8-24.
Figure 8-24 Daily Schedule Configurator
The configurator page contains a number of controls that define the schedule for generating the report daily:
•
•
•
Click OK to save the schedule, or Cancel to cancel it.
Weekly Schedule
A Weekly schedule is for generating the report on one or more days each week. The weekly configurator page is shown in Figure 8-25.
Figure 8-25 Weekly Schedule Configurator
The configurator page contains a number of controls that define the schedule for generating the report weekly:
•
•
•
•
Click OK to save the schedule, or Cancel to cancel it.
Monthly Schedule
A Monthly schedule is for generating the report on one day each month. The monthly configurator page is shown in Figure 8-26.
Figure 8-26 Monthly Schedule Configurator
The configurator page contains a number of controls that define the schedule for generating the report monthly:
•
•
•
•
•
Click OK to save the schedule, or Cancel to cancel it.
Yearly Schedule
A Yearly schedule is for generating the report on one day each year. The yearly configurator page is shown in Figure 8-27.
Figure 8-27 Yearly Schedule Configurator
The configurator page contains a number of controls that define the schedule for generating the report yearly:
•
•
•
•
•
Click OK to save the schedule, or Cancel to cancel it.
Editing a Schedule
You can edit the schedule for a saved report by clicking the Edit link next to the schedule. The schedule configurator page is displayed, showing the original schedule settings.
You can change the schedule settings and click OK to save the modified schedule, or click Cancel to keep the original settings.
For details on using the schedule configurator for each type of schedule, see the "Adding a Schedule" section.
Deleting a Schedule
You can delete a schedule for a report by clicking the Delete link next to the schedule. Note that there is no confirmation before the schedule is deleted.
Configuring Email for Scheduling
The email configuration for scheduled reports must be set up first before you can schedule any reports. To configure email settings, click the Email Configuration link at the top of the Scheduled Reports page. This displays the Email Configuration page, shown in Figure 8-28.
Figure 8-28 Email Configuration Page
The mail configuration page defines the email account settings that are to be used for distributing scheduled reports by email. It includes the following fields:
•
•
•
•
•
•
•
•
•
•
•
Click OK to save the email settings, or Cancel to cancel the operation.
Managing Locations
Use the Manage Locations command to define and manage locations. A location is simply a named range of source IP addresses. Locations can be used in AppScope reports to easily group results from particular IP address blocks.
Click this command to display a page showing a list of defined locations, where you can delete existing locations and define new ones, as shown in Figure 8-29.
Figure 8-29 Locations Page
To delete a location, check the Delete checkbox to the left of a location (or click the top checkbox to delete all locations) and click Delete Checked Locations.
To add a new location, click the Add Location link. This displays a form where you specify the location name and the beginning and ending IP addresses of the corresponding range, as shown in Figure 8-30. Click the Save button to save the new location.
Figure 8-30 Add Location Page
To edit or delete a location, click on the name of the location in the list (as shown in Figure 8-29). The Edit Locations page shown in Figure 8-31 is displayed.
To edit the location, you can do any of the following:
•
•
•
•
Click Save to save your changes, or click Delete to delete the entire location including all ranges it defines.
Figure 8-31 Edit Locations Page
Defining Transaction Types
A transaction type is represented by a Boolean expression that contains criteria that determine whether a given transaction matches this type or not. You can use transaction types in AppScope reports by limiting the report to a single transaction type with the Transaction Type pull-down list on the AppScope Reports Query page.
An example of a simple expression is this:
Domain Matches hrportal.company.com
A more complex expression is this:
Domain matches hrportal.company.com
And
URL matches .*/submitForm.jsp$
And
Not ParamSummary[_test] Exists
And
(
ParamSummary[RequestType] matches EmploymentReq
Or
ParamSummary[RequestType] matches PerfReview
)Use the Transaction Types command (under the Reports folder in the left pane) to view a list of all defined transaction types, as shown in Figure 8-32.
Figure 8-32 Transaction Types Page
Each row in the table lists one saved transaction type; the transaction type name is shown, followed by its sequence number and definition. Sequence numbers are explained in the next section, Adding a Transaction Type. To change a sequence number, enter a new number in the sequence column and click Update Sequence.
To add a transaction type, click the Create Transaction Type link above the table. Click the Edit or Delete links next to a transaction type to edit or delete the transaction type.
You can group transaction types into logical groups for reporting purposes through a NMS (Network Management System). For details, see the "Transaction Grouping" section.
Adding a Transaction Type
You can add a transaction type by clicking the Create Transaction Type link above the table. The create/edit transaction type page is displayed, as shown in Figure 8-33.
Figure 8-33 Create/Edit Transaction Type Page
Enter a name for the transaction type in the Name field.
In the Definition area, you can click New to create a new expression for the transaction type (for details, see the "Adding a New Expression" section). If there are existing expressions, the new expression is added after the row in which you click New. The new expression is joined to the existing expression with an And connector. You can change this (or any connector between two expressions) to an Or connector by choosing Or in the pull-down list.
If you have multiple expression lines, you can add parenthesis to group lines by clicking the checkbox for the first and last lines you want to enclose in parentheses, and then clicking the Go button in the Add Parentheses section.
To delete parentheses, click Delete on the row of either the opening or closing parenthesis. The matching parenthesis is automatically deleted.
To delete an expression, click Delete on the row containing the expression. A confirmation dialog will ask you to confirm the deletion. If you delete an expression that is the only thing enclosed in a set of parentheses, then the set of parentheses is also deleted.
The Substitute Client Timing section is explained in the next subsection, "Substitute Client Timing".
To save the transaction type, click Save, or to cancel its creation or editing, click Cancel.
Substitute Client Timing
In parts of some web applications, the returned page is not rendered by the browser, but is used by some other client side component such as a Java applet or ActiveX control. For such a request, the server time can be measured, but not the client-side timings. Usually AppScope does not report on requests where no client-side timings are received, because the assumption is the request was not completed or encountered an error.
Check the Use Substitute Client Timing checkbox to enable AppScope to measure at least the server timing for such requests. In such cases, AppScope uses 0 for all client-side times, so the reported Server Time, TTFB, TTLB, and Total Time will all be the same as the Server Time. To use more realistic values for the client-side times, you can enter the time in milliseconds for the average network latency (one way) and client rendering time. If left blank, these fields default to 0.
When a transaction is measured, it may match more than one defined transaction type. If substitute client timing is enabled for the transaction types, there needs to be a way to specify which substitute timing statistics to use, since these may be different for different transaction types. The sequence numbers on the Transaction Types page (Figure 8-32) determine the transaction type that is used for timing purposes. The first matching transaction type (the one with the lowest sequence number) is used for substitute timings.
As each transaction type is created, it is given an incremented sequence number. These numbers can be viewed and changed in the Sequence column on the Transaction Types page (Figure 8-32).
The Substitute Client Timing feature can be used in combination with the AppScopeLogNonInstrumented Application Class keyword to allow AppScope to report on requests where the response is not HTML or cannot be instrumented for measurement. For more details, see Table 5-2.
Adding a New Expression
When you click New to create a new expression, or Edit to edit an existing expression, the create/edit expression page is displayed, as shown in Figure 8-34.
Figure 8-34 Create/Edit Expression Page
This page contains the following controls:
•
•
•
•
If you select ParamSummary in the Attribute list, the form changes to add a Param Name field, as shown in Figure 8-35.
Figure 8-35 Create/Edit Expression Page with ParamSummary
For the ParamSummary attribute, set these other fields:
•
•
When you are finished defining or editing the transaction expression, click OK to save it and return to the create/edit transaction type page, or click Cancel to cancel creating or editing the expression.
Table 8-2 Transaction Type Attributes
Domain
The server name, for example, www.example.com
URL
The URL, not including the query string, if any
QueryString
The first 255 characters of the query string
ParamSummary
A summary of all the query parameters in the request, including the full text of each parameter name and the first 100 bytes of each parameter value. (The number of bytes is configurable by the ParamSummaryParamValueLimit keyword in the fgn.conf file; see Table 5-2.)
The ParamSummary attribute is indexed by the parameter name; for example, the expression ParamSummary[UserAction] returns the value of the UserAction parameter.
The parameter can be very large in the case of a browser POST, so the scan for parameters is limited to the first 40 KB of posted data, be default. You can change this by using the PostContentBufferLimit keyword in the fgn.conf file; see Table 5-2.
The ParamSummary attribute is not available by default and must be enabled by the ParamSummary keyword in the fgn.conf file; see Table 5-2.
ApplicationClass
Allows the Application Class of a request to be used to determine its transaction type.
RequestGroupingString
Allows the request grouping string of a request to be used to determine its transaction type. The request grouping string is defined by the RequestGroupingString keyword; see Table 5-1.
The ParamSummary and QueryString attributes contain similar data, but each has its advantages and disadvantages:
•
•
•
•
Editing a Transaction Type
You can edit a transaction type by clicking the Edit link on the same row as the expression. The create/edit transaction type page is displayed, as shown in Figure 8-33. The process for editing a transaction type is exactly the same as described in the "Adding a Transaction Type" section.
Deleting a Transaction Type
You can delete a transaction type by clicking the Delete link on the same row as the expression. A confirmation dialog will ask you to confirm the deletion.
You cannot delete a transaction type that is used in a business transaction unless it is removed from the business transaction or the business transaction is deleted.
Managing Transaction Type Mapping
At the time each transaction is recorded in the database, AppScope determines its transaction type(s) and stores them in the database. This is known as the mapping between transactions and defined transaction types. When a query is executed using the transaction type filter, the stored transaction type mappings are used to determine which results to display. For information on using transaction types in queries, see the Transaction Type pull-down list in the "Defining the Query" section.
Because you can update the transaction type definitions at any time, some of the mappings in the database may become "stale." There are two ways to deal with this:
•
•
To reach the Transaction Type Mapping Status page, use the Transaction Types command (under the Reports folder in the left pane). Click on the Transaction Type Mapping Status link in the upper left of the page to display the page shown in Figure 8-36.
Figure 8-36 Transaction Type Mapping Status Page
In the above example, you can see that the current transaction type version number is 61. There are a total of 34490 transactions in the database. None of the transaction type mappings are current because their transaction type version is 41 or 0.
You can update the mappings for some or all of the transactions by specifying a transaction date and clicking Update. In our example, if we update all of the transaction mappings we will get a result screen like that shown in Figure 8-37. Notice that because the transaction type mappings are current, there is no Update button available.
Figure 8-37 Transaction Type Mapping Status Updated Page
Defining Business Transactions
Business transactions are a way of grouping transactions into sets. For example, imagine that a user visits five web pages to file an online expense report. You can group this set of five transactions into a single business transaction. You can run AppScope reports on business transactions by setting the Organize Output By pull-down list to "Business Transaction" on the AppScope Reports Query page.
Note that a business transaction does not track the flow of a given user through a web application. Instead, individual transaction data is collected and aggregated into statistics based on the frequency that each transaction occurs in the typical sequence. Business transaction data is aggregated, not gathered on a per-user tracking basis.
To define a business transaction, follow these steps:
1.
2.
3.
Use the Business Transactions command (under the Reports folder in the left pane) to view a list of all defined business transactions, as shown in Figure 8-38.
Figure 8-38 Business Transactions Page
Each row in the table lists one saved business transaction; the name is shown, followed by its component transaction types. You can click on a transaction type name to jump to the Transaction Type page where it is defined.
To add a business transaction, click the Create Business Transaction Type link above the table. Click the Edit or Delete links next to a business transaction to edit or delete the business transaction.
•
•
•
Adding a Business Transaction
You can add a business transaction by clicking the Create Business Transaction Type link above the table. The create/edit business transaction page is displayed, as shown in Figure 8-39.
Figure 8-39 Create/Edit Business Transaction Page
In the Available Transaction Types area, select the transaction types that you want to include in the business transaction. Click to select a single entry in the list, Shift-click to select a range of entries, or Ctrl-click to select disjoint entries. You can also use the Select All or Select None buttons. When you have selected the desired transaction types, click Add Selected Types to Business Transaction. (The figure shows that two transaction types have already been added to the business transaction.)
In the Business Transaction Definition area, enter a name for the business transaction in the Name field. This name must be unique among both business transactions and transaction types.
In the Component Transaction Frequency area, select the appropriate radio button for how the frequency of a transaction is to be computed:
•
•
–
–
The Component Transaction Types area allows you to enter frequency information and reorder the transactions.
If you elected to use specified frequencies, enter the number of times each individual transaction occurs in the business transaction. If you elected to use actual frequencies, these fields are ignored.
Click the Counter radio button for the transaction that you want to use as the "counter" transaction. When this transaction occurs, then an instance of the business transaction is considered to have occurred.
The Select checkboxes are for selecting a transaction to move up or down in the list, which you do by clicking Move Up or Move Down. Note that the sequence of transactions has no effect on statistics and the ability to change the sequence is provided simply to improve readability and maintenance. You can also select one or more transactions and click Remove to remove them from this business transaction.
When you are done creating or editing the business transaction, click Save to save it. If there is an error, you are returned to this page, with an error message at the top explaining what is wrong.
Editing a Business Transaction
You can edit a business transaction by clicking the Edit link on the same row as the business transaction in the list. The create/edit transaction type page is displayed, as shown in Figure 8-39. The process for editing a business transaction is exactly the same as described in the "Adding a Business Transaction" section.
Deleting a Business Transaction
You can delete a business transaction by clicking the Delete link on the same row as the business transaction in the list. A confirmation dialog will ask you to confirm the deletion.
Upload Data
Use the Upload Data command to manually upload all log data from nodes to the Management Console database. Normally, you don't need to use this command because log data is automatically uploaded every 30 seconds. However, if you are testing, you can use this command to force an immediate upload of data since the last upload.
Figure 8-40 Upload Data Page
After you click Yes, the Management Console displays a status page showing the number of records that were uploaded, and the servers involved.
