Specifying Options for Reports

Strobe accepts processing options first from the parameter field of the EXEC statement that invokes a Strobe procedure. It then accepts options from a SYSIN data set or the OTHER PARAMETERS field in the Strobe - Detail For A Performance Profile panel in Strobe/ISPF. If the same parameter is found more than once, it uses the last one specified, except for COMPRES and DETAIL, where all values are concatenated. If you specify a parameter, but do not specify a value, Strobe uses the default value. The parameters are identified by keywords.

Providing a Report Subtitle

You can define REPNAME to specify a subtitle that will appear on the title line of the Performance Profile. REPNAME defaults to the name of the target program. The report subtitle cannot exceed 40 characters. If the name contains commas, spaces, or other special characters, enclose it in apostrophes or parentheses. Use parentheses within apostrophes when the value contains special characters and is specified as an EXEC parameter; for example:

Specifying Report Resolution

You can set RESLTN (default 64) to break up the report into manageable sections. RESLTN specifies the report resolution—the number of bytes to be considered a codeblock for detailed reporting on the Program Usage by Procedure report. Enter a value between 2 and 9999. For indexed source support, Strobe uses this value to subdivide large procedures into codeblocks. Strobe combines consecutive source language statements within a procedure until their size approximates the specified report resolution. For unindexed modules, each codeblock is defined as a block of machine code equal in length to the report resolution. This option is ignored for iStrobe format profiles.

Specifying Sort Core Size

SORTSIZ (default 999999) specifies the sort core size in bytes that is available to the system sort/merge program invoked by Strobe. Enter a value of 0 or from 12000 to 999999. To allow the sort/merge program access to all available space in the region, omit this parameter. To instruct the sort/merge program to use its system default, enter a value of 0; for example:

Specifying the Page Size of the Performance Profile

LINEMAX (default 60) specifies the number of lines that appears on each page of the Performance Profile. Change this value if your installation uses a short paper stock or prints eight lines per inch. Use a value from 45 to 80; for example:

Specifying the Date and Time Formats on the Performance Profile

There are four parameters that control the date and time formats printed in various reports in the Performance Profile: DATE, DATESEP, TIME, and TIMESEP.

DATE specifies the date format printed in various reports in the Performance Profile. The format options are:

DATESEP specifies the date separator printed in various reports in the Performance Profile. The separator options are:

TIME specifies the time format printed in various reports in the Performance Profile. The format options are:

TIMESEP specifies the time separator printed in various reports in the Performance Profile. The separator options are:

Obtaining Detailed Reporting for System Modules

DETAIL allows you to provide a list of system modules or control sections that Strobe normally treats as pseudo-entities. (See the discussion on pseudo-entities in Strobe Interpretation and Analysis User Guide.) To obtain detailed Program Usage by Procedure subreports for operating system modules, supply the module names after the DETAIL keyword. These operating system modules will then appear as user modules in Strobe reports. If you specify more than one name, use a comma-delimited, parenthetical list. You can specify as many as 50 names. For example,

Specifying Modules to Be Treated as Pseudo-Sections

COMPRES allows you to combine modules with identical initial characters into your own named pseudo-sections on the Program Section Usage Summary and the Program Usage by Procedure reports. The pseudo-sections that you specify are added to those provided by Strobe. You can specify as many as 25 pseudo-sections to compress.

In parentheses, specify the prefix, followed by an equal sign (=), and your pseudo-section name. Each prefix can contain as many as eight characters, and each pseudo-section name can contain as many as seven characters preceded by a period (.). Specifying the full eight- character prefix compresses one module, and only that module, into a pseudo-section.

For example, if the common prefix is three characters long, specify:

COMPRES=(IKJ=.TSO)

To compress more than one group, use a second set of parentheses:

COMPRES=((IKJ=.TSO),(IGX00=.RMF))

To place modules with different prefixes into the same pseudo-section, specify them separately:

To compress one module into a pseudo-section, specify:

Specifying the Suppression of Mapping for Modules

NOMAP enables you to specify that module mapping will not occur for any module. To suppress module mapping for all modules, specify:

For example, to suppress module mapping for a module named MOD1, specify:

To suppress module mapping for more than one module (to a maximum of 10), use a comma-delimited, parenthesized list. For example:

Tailoring the Program Usage by Procedure Report

NOPROC suppresses the detailed Program Usage by Procedure report. Specify:

Specify PUBP (default 0) to specify a minimum percentage of CPU activity that a procedure code block must exhibit for Strobe to produce a line for it in the Program Usage by Procedure report, specify:

All contiguous procedures that exhibit less than this percentage are condensed on a single line in the Program Usage by Procedure report. You can specify a value anywhere between 0 and 99.9. For example to specify a minimum CPU activity of 10%, specify:

If the result for a control section would be only one report line, Strobe suppresses reporting for that control section.

Tailoring the DASD Usage by Cylinder Report

There are two parameters you can specify to condense the DASD Usage by Cylinder report: NODASD|DASD and DASDGAP=nn.

NODASD suppresses the detailed DASD Usage by Cylinder report. Specify:

DASD=nn.n (default 2) condenses information about all contiguous cylinders that exhibit less than the specified percentage of DASD usage into a single line in the DASD Usage by Cylinder report. Specify a value from 0 to 99.9. For example, to set a minimum of 25 percent of DASD usage, specify:

Used with the DASD= parameter, DASDGAP=nn (default 5) specifies the number of inactive cylinders that can be passed over while Strobe continues to accrue DASD usage. These cylinders are then presented as a single line on the DASD Usage by Cylinder report. Specify a value from 0 to 99.

Consider the following example. You specify DASD=10 and DASDGAP=4. During the measurement session, Strobe identifies activity for cylinder numbers 103 through 107, and then for cylinder numbers 109 through 113; however, all of the identified activity is below the specified minimum of 10%. Strobe presents this information as a single line in the DASD Usage by Cylinder report, even though there is a gap between the cylinder ranges.

For example, to specify that 10 contiguous cylinders without identified activity can be passed over, specify:

Tailoring the Attribution Reports

NOATTR suppresses all the Attribution reports. Specify:

NOATTR=attrtype suppresses all the Attribution reports for the specified type of attribution. You can suppress reports for SVCs, Db2, CICS, CA-IDMS, CA Gen, PL/I, COBOL, C, IMS, WebSphere MQ, JAVA, and VSAM. (To suppress CA Gen attribution reports, use the IEF keyword.)

For example, to suppress Attribution reports for Db2, specify:

To suppress reports for more than one type of attribution, use a comma-delimited, parenthesized list. For example, to suppress Attribution reports for SVCs, Db2 and COBOL, specify:

ATTR=nn.n (default 2) suppresses Attribution reports for any system service modules in which the total CPU or wait time percentage is less than a specified baseline percentage. Specify a value from 0 to 99.9. For example, to suppress Attribution reports for system service modules that show less than 9.9 percent of the total CPU or wait time, specify:

ATTRLINE=nn.n (default 0) specifies the minimum percentage of CPU time or wait time that a module/transaction must exhibit to produce a separate line in the Attribution reports. Specify a value from 0 to 99.9. For example, to show only modules or transactions that exhibit more than 25 percent of CPU or wait time on a single line, specify:

Tailoring the Transaction Usage by Control Section Report

NOTRAN suppresses the detailed Transaction Usage by Control Section report. For example, specify:

TUCS=nn.n (default 0) specifies the minimum percentage of CPU activity that a transaction must exhibit to produce a report within the Transaction Usage by Control Section report. Specify a value from 0 to 99.9. For example, to limit transactions appearing in the report to those showing at least 49.9 percent of CPU activity, specify:

Tailoring the Time and Resource Demand Distribution Reports

NOTASK combines the execution activity for all tasks into one execution line in the Time and Resource Demand Distribution reports. For example, specify:

ALLTASK shows all tasks, including those that show no activity, in the Time and Resource Demand Distribution reports. The default is for Strobe to show only the six tasks that display the highest execution activity. To show all tasks, specify:

ALLDD shows all dd statements, including those that show no activity, in the Time and Resource Demand Distribution reports. The default is for Strobe to show only the 11 DD statements with the highest activity. To show all DD statements, specify:

Expanding the Program Section Usage Summary Report

ALLCSECT shows all control sections, including those that show no activity, in the Program Section Usage Summary report. The default is for Strobe to show only active control sections. To include control sections that show no activity, specify:

Producing the Most Extensively Inactive Storage Areas Report

The Most Extensively Inactive Storage Areas report is suppressed by default, unless you specify MEISA. To produce the report, specify:

Expanding the Wait Time by Module Report

Specifying WAITLOC causes the location of program wait in each module to be shown in the Wait Time by Module report. To show the wait time, specify:

Suppressing the Data Set Characteristics Supplement Report

NODSCS suppresses the Data Set Characteristics Supplement report. To suppress the report, specify:

Suppressing the Coupling Facility Activity Report

CF=YES|NO controls the production of the Coupling Facility Activity report. Specify YES to produce the Coupling Facility Activity report if there is coupling facility activity on the system. Specify NO to suppresses production of this report. If you do not specify the CF parameter, Strobe produces the report if it detects activity in IXL modules.

Suppressing the z/OS Memory Objects Report

NOMEMOBJ suppresses the z/OS Memory Objects report. To suppress the report, specify:

Tailoring the Reporting of CA Gen Activity

IEF enables you to specify how you want the Performance Profile to present CA Gen activity in the CPU Usage by IEF Statement report and in the Wait Time by IEF Statement report. The following are the keywords for the IEF parameter:

TIMEPCT=nn.n (default 0) is a value from 0.0 to 99.9 representing the minimum percentage of CPU or run time for a CA Gen statement detail line to be printed on the report. The higher the percentage you specify, the more concise the report. The lower the percentage, the more detailed the report. For example, to print detail lines that display activity above 15%, specify:

TOPTEXT=nn (default 3) is a value from 0 to 99 representing the number of CA Gen statements with the most identified CPU activity for which you want to see full statement text. A value of 99 shows full statement text for all CA Gen statements. For example, to print the statement text for the five statements that consumed the most CPU, specify:

OFFSET specifies that activity is reported by offset within each COBOL statement within SQL statement within CA Gen statement number. (The default is to suppress the offset.) To show activity by offset, specify:

NOSQL suppresses the display of SQL statement numbers associated with CA Gen statements. Specifying NOSQL results in a more concise report. (The default is to display the SQL statement numbers.)

To suppress the display of SQL statement numbers, specify:

Tailoring the Reporting of SQL Activity

The SQL parameter enables you to specify how you want the CPU Usage by SQL Statement and Wait by SQL Statement reports present SQL activity.

SQL=CPUTEXT=nn.n (default 10.0) enables you to specify the SQL statements for which full text will be printed. nn.n can range from 0.0 to 99.9 representing the minimum percentage of CPU or run time for full SQL statement text, regardless of its length, to be printed in the CPU Usage by SQL Statement report. The higher the percentage you specify, the more concise the report. The lower the percentage, the more detailed the report. SQL statements that do not meet the specified percentage default to the length specified in the MAXLEN parameter described previously. For example, to print full SQL text for statements that display activity above 15%, specify:

SQL=WAITTEXT=nn.n (default 5.0) enables you to specify the SQL statements for which full text will be printed. nn.n can range from 0.0 to 99.9 representing the minimum percentage of wait time for full SQL statement text, regardless of its length, to be printed in the Wait by SQL Statement report. The higher the percentage you specify, the more concise the report. The lower the percentage, the more detailed the report. SQL statements that do not meet the specified percentage default to the length specified in the MAXLEN parameter described next. For example, to print detail lines that show wait above 25%, specify:

SQL=MAXLEN=nnnn enables you to control the length of the SQL text printed in the CPU Usage by SQL Statement and the Wait by SQL Statement reports. Enter a value that is a multiple of 100 from 100 to 9900 (default 300). This value represents the maximum number of characters printed on the report for those statements that do not meet the threshold defined in the CPUTEXT or WAITTEXT parameters. To print all the SQL text regardless of its length, enter 9999. Values that are not multiples of 100 are rounded up to the next multiple of 100. To suppress all SQL text, enter 0. For example, to print the entire SQL text for each Db2 query, specify:

Suppressing the DL/I CPU and Wait Reports

NODLI suppresses the DL/I CPU and Wait reports, for installations with Strobe for IMS. To suppress the reports, specify

Tailoring the Reporting of CICS Transaction Profiling Information

The following CICS keywords enable you to tailor how Strobe reports data for the CICS transactions and regions for which Strobe has gathered detailed information.

NOTXPRPT enables you to suppress the following CICS-related reports: Transaction Summary, Transaction Profile, and the Region Level.

NORLIRPT enables you to suppress CICS-region level reports.When you specify this parameter, Strobe does not produce the Region Suspend by Class and Region Suspend by Resource within Class reports.

TXPONLY enables you to suppress all Performance Profile reports except the Measurement Session Data, the Token - Longname Cross Reference, and all CICS Transaction Profile reports. When you specify this keyword, Strobe produces the following reports: Transaction Summary, the Transaction Profile, the Region Suspend by Class, the Region Suspend by Resource within Class, the Measurement Session Data, and the Token - Longname Cross Reference reports. If you also specify NORLIRPT, Strobe additionally suppresses the CICS Region Level reports.

TLIBASE=nn.nnn (default 00.005) enables you to control the level of detail in the following Transaction Profile reports by specifying a minimum value in seconds.

• On the CICS API Service Time report, Strobe reports the Exec and Suspend Mean Service times that are less than this value under the pseudo-section .COMPRES.
• On the Non-API Service Time report, Strobe reports under the pseudo-section .COMPRES all Resource Names, Resource Types, modules and control sections that have Mean Service times that are less than this value.
• On the API Service Time Detail report, Strobe suppresses all Resource Type and Resource Name combinations with Mean Service Time less than the time specified. Strobe does not report this activity under a pseudo-section. Similarly, Strobe suppresses all modules and control sections with Exec Mean Service Time that is less than this value. These modules and control sections do not appear on the report. If the detail lines for a command do not meet or exceed this value, Strobe suppresses the command. If Strobe suppresses all commands, Strobe does not produce the report.

RLIBASE=nn.nnn (default 00.005) enables you to control the level of detail in the Region Suspend by Class and the Region Suspend by Resource within Class reports by specifying a minimum value in seconds.

• On the Region Suspend by Class report, Strobe reports under the pseudo-section .COMPRES all Suspend Classes with Suspend Time for an Average Transaction that is less than this value.
• On the Region Suspend by Resource within Class report, Strobe suppresses a Resource with Suspend Time for an Average Transaction that is less than this value. This resource does not appear on the report. If Strobe suppresses all Resources for a Class, Strobe does not produce the report for that class.

Compressing WebSphere MQ Information

The MQCALL=nn.n (0) variable enables you to condense activity reported for WebSphere MQ calls on the CPU Usage by Module by MQSeries Call and the Wait by Module by MQSeries Call reports. The value nn.n stipulates the minimum percentage of activity for a WebSphere MQ call at a given location in which Strobe detects activity. Strobe suppresses reporting of all activity below this percentage.

Obtaining IMS Transaction-Level Information for WebSphere MQ Calls

The MQTRAN parameter enables you to obtain IMS transaction-level information, in addition to WebSphere MQ module-level information, on the CPU Usage by Module by MQSeries Call and the Wait by Module by MQSeries Call reports.

Specifying Parameters for a Temporary Work File for iStrobe

If iStrobe is installed at your site, you can use the following parameters to specify values for a temporary work file:

• ALOCUNIT=type enables you to determine the type of unit: BLK (block), CYL (cylinder), or TRK (track). The default is BLK.
• WRKUNIT=unit enables you to specify the unit for the temporary work file. The default is SYSDA.
• WRKVOL=volume enables you to specify the volume for the temporary work file.
• WRKPRIM=nnnn enables you to specify the primary space allocation for the temporary work file. For the following ALOCUNIT specifications, these are the defaults: BLK, 600; CYL, 8; TRK, 120.
• WRKSEC=nnnn enables you to specify the secondary space allocation for the temporary work file. For the following ALOCUNIT specifications, these are the defaults: BLK, 600; TRK, 120; CYL, 8.

For the WRKPRIM and WRKSEC parameters, if the value you specify is less than the default value, Strobe uses the default value.

Collecting and Customizing AutoStrobe History Records

If you have AutoStrobe installed at your site, the following parameters enable you to collect and customize measurement session history records.

COLLHIST requests the creation of a measurement session history record for this measurement session. Once the record is created, you can examine the measurement session history for a specific job step, DBRM, or a transaction. To request the creation of a measurement session history record, specify:

CMT=”dddd” adds a comment to a measurement history record, if one does not already exist. The character string, enclosed in double quotes, that follows CMT= represents the comment. Comments have a maximum length of 55 bytes. Embedded double quotes and underscores are not allowed in the character string.

To enter a comment for a measurement session history record, specify:

DBRMBASE and TRANBASE enable you to control the amount of information stored in the measurement session history record.

DBRMBASE=nn.n specifies a baseline percentage of CPU execution samples a DBRM must exhibit for Strobe to save measurement session history record for that DBRM. If the DBRM has less than the specified baseline percentage, it is not included in the measurement session history. The default baseline is 1.0. To enter a DBRM baseline of 10%, for example, specify:

TRANBASE=nn.n specifies a baseline percentage of CPU execution samples a transaction must exhibit for Strobe to save measurement session history record for that transaction. If the transaction has less than the specified baseline percentage, it is not included in the measurement session history. The default baseline is 1.0. To enter a DBRM baseline of 8.5%, for example, specify:

REPJOB=jobname specifies the replacement name for the job name in the measurement session history record. Use this parameter to override the actual job name with a substitute job name. This parameter enables you to compare jobs that are the same program execution, but may have different job names. The job name can be up to eight characters in length, but cannot include an asterisk (*) or embedded blanks.

REPSTP=stepname specifies the replacement name for the step name in the measurement session history record. Use this parameter to override the actual step name with a substitute step name. This parameter enables you to compare job steps that execute the same program, but may have different step names. The step name can be up to 17 characters in length, but cannot include asterisks (*) or embedded blanks.