The R03 Schedule New Measurement dialog always starts with panel
1 (Job Information) selected, it is shown here:
File View Navigate Help
--------------------------------------------------------------------------------
R03: Schedule New Measurement Row 00001 of 00019
Command ===> __________________________________________________ Scroll ===> CSR
1. Job Information 3. Multi Steps 5. Subsystems 7. Schedule
2. Options 4. Active Jobs 6. Sysplex 8. Sched Options
________________________________________________________________________________
Panel 1. Job Information
Job Name/Pattern . . ________
(Inactive)
Step Specification
Step No. . . . . . ___ Specify step number, program name,
Program Name . . . ________ step name or step name + Proc step
Step Name . . . . ________ name. Use panel 3 to specify more
ProcStepName . . . ________ than one step.
Description . . . . __________________________________________
Number of Samples . _______ Measure to step end . . . N
Duration (min:sec) . _______ Delay by (secs) . . . . . ___
Notify TSO User . . _______ Retain file for (days) . 20
USS observations . . . . Max. 20
The panel shown above was invoked with a NEW primary command, so
input fields are mostly blank.
Panel 1 - input fields
- Job Name/Pattern
- This field is mandatory. It cannot be blank. Specify the name
of the job, started task, or TSO region to be measured.
- Searching for active jobs (*)
- You can specify a job name pattern; for example, a
job name prefix followed by an asterisk (*), a job name suffix preceded
by an asterisk, or an asterisk by itself. The asterisk indicates that
the region to be measured is currently active. A list of active jobs
whose names match the wildcard pattern will be displayed in panel
4, where one or more can be selected for measurement.
- Creating multi-job measurements (%)
- You can specify wildcards in order to generate measurements
for multiple jobs in one request. To sample multiple jobs in one request,
use a percent (%) as a wild card character anywhere in the job name.
Multiple percent characters may be used in the job name, such as @(#)
11 1.1@(#)B% to sample any jobs with A and B in the job name separated
by 0 to many characters. Panel 4 can be used to display the currently
active jobs that will be sampled.
A job name pattern is only permitted
in NEW or MOD requests. It is not valid for Threshold or Trigger requests.
You can not specify a jobname pattern of "%". If a pattern
is present in the jobname, then the request will automatically be
set to only select active jobs.
The multiple job feature only
works with jobs that are active. It does not wait for jobs to start.
You can schedule a multiple job request to start at a future date
and time by using panel 7 Schedule, but only one future event is permitted.
The jobs that are active at the scheduled date and time (and matching
the wildcard criteria) will be measured.
The maximum number
of jobs that can be measured from a multi-job request is defined during
the installation of Application Performance Analyzer. When this limit is exceeded, Application Performance Analyzer stops
creating measurements for this request and the status of the request
is displayed as ‘Stoppd'. The measurements that executed
(within the limit) are accessible for report viewing under the request.
To increase the limit, contact your system programmer.
- Obtaining DB2® DDF Data
- To obtain DDF data, you must measure the DB2 DDF
address space (xxxxDIST) with the DB2+ extractor turned
on. You have the option of limiting the scope of a DDF measurement
by specifying filtering criteria in Panel 5 Subsystems. DDF measurements
may be filtered by Correlation Id, End User Id and/or Workstation
Id. For more information on measuring DDF activity, refer to Measuring DDF activity.
- Measuring a specific DB2 stored procedure or user-defined
function (-)
- To measure a specific DB2 stored procedure or user-defined function,
use a dash (-) as the first and only character in the Jobname/Pattern
field. The information identifying the DB2 stored
procedure or user-defined function must be entered on Panel 5. Subsystems.
This feature is only available when the WLM Intercept is activated
during Application Performance Analyzer installation, and you are given appropriate
security access to it. Contact your system programmer for access if
necessary. When measuring a specific DB2 stored
procedure or user-defined function, the following measurement options
are not applicable and will result in an error if used: Step specification/Multi
Steps, Measure to step end, Delay by, Active Jobs, CICS and IMS selection
criteria, Schedule and Schedule Options.
- Measuring a specific IMS transaction across
multiple MPP regions
- To measure a specific IMS transaction that executes in multiple
MPP regions, you must enter a dash (-) as the first and only character
in the Jobname/Pattern field. The IMS transaction name and the IMS
subsystem name or IMSPLEX group name must be entered on Panel 5 Subsystems.
You must also select the MPP regions you want to monitor in Panel
4. Active Jobs. This feature is only available when the IMS Intercept
is activated during Application Performance Analyzer installation,
and you are given appropriate security access to it. Contact your
system programmer for access if necessary. When measuring a specific
IMS transaction across multiple MPP regions, the following measurement
options are not applicable and are ignored if used: Step specification/Multi
Steps, Measure to step end, Delay by, Schedule and Schedule Options.
- Active/Inactive indicator
- When a NEW command is entered, Application Performance Analyzer checks
for and displays the current status of the job, started task, or TSO
region immediately below the jobname. When it is detected as active, ‘Active'
is displayed, otherwise, ‘Inactive' is displayed. It is only
necessary to use Panel 8 Sched Options if you wish to change the observation
status from that detected by Application Performance Analyzer.
When more
than one active job is selected for measurement in Panel 4, the phrase ‘(Active
- Multiple Jobs Selected)' is displayed below the Jobname. In
this case, the name listed is the first job selected in Panel 4. You
must use Panel 4 to view or change the jobs that are selected for
measurement.
- System name
- This field appears only if the Application Performance Analyzer you are connected
to is configured as a member in a SYSPLEX group. Specify the name
of the system on which the measured job is to run (or is currently
running). Specify an asterisk (*) in this field to indicate that the
job could run on any of the systems in the group.
You can also
select panel 6 to display a full list of available system names from
which you can make a selection.
- Step Specification Field Group
- Step specification is not applicable when measuring specific DB2 stored
procedures or user-defined functions.
The job step to be measured
is specified by a group of four fields (Step Number, Step Program,
Step Name, ProcStepName). To identify the step, you can specify one
of the following:
- Step Number by itself
- Step Program by itself
- Step Name by itself
- Step Name and ProcStepName
If you leave all of these fields blank, the first job step
is assumed. You cannot specify any of these fields when you are specifying
measurement of a job that is currently active.
To schedule the
measurement of multiple steps in the same job, select panel 3.
- Step No.
- If Step No. is specified, the other three step fields (Step
Program, Step Name, and ProcStepName) must be left blank.
Step
No. specifies the numeric step number.
For
a NEW measurement, you may enter an asterisk (*) in this field to
measure all steps in the job.
For a threshold
(TNEW) measurement, you may enter an asterisk (*) in this field to
measure all steps in the job that meet the threshold criteria entered
in the Criteria panel.
- Step Program
- If Step Program is specified, the other three step fields (Step
No., Step Name, and ProcStepName) must be left blank. Program Step
specifies the name of the program coded in the PGM= parameter of the
EXEC statement for the step you want to measure.
- Step Name
- If Step Name is specified, then Step Number, and Step Program
must be left blank. You can specify Step Name by itself or in combination
with ProcStepName.
Step Name specifies the symbol coded in the name
field of an EXEC PGM = statement or an EXEC PROC = statement. In the
event that the ProcStep name field is also supplied, this field always
identifies the symbol coded in an EXEC PROC = statement.
In
the event that the ProcStep name field is left blank, and Step Name
matches an EXEC PROC = statement, the first step within that proc
will be measured.
If the step to be measured is not within a
proc, then Step Name specifies the symbol coded in the name field
of an EXEC PGM = statement, and ProcStepName must not be specified.
- ProcStepName
- If ProcStepName is specified, then input must also be supplied
in the Step Name field.
ProcStepName specifies the symbol coded
in the name field of an EXEC PGM = statement that is part of a PROC.
- Description
- Enter a description for this observation request. This field
is optional unless the option has been set during installation requiring
a minimum description of 8 characters.
- Number of Samples
- Specify the number of times execution of the measured jobstep
is to be sampled. Samples are taken in equal intervals. The sampling
frequency is determined by dividing the number of samples by the specified
measurement duration. If configured during installation, a default
value will be displayed in this field for NEW and TNEW requests. An
installation default value is used if you do not supply input in this
field.
Note:
This field is not used when sampling
a DB2 DDF address space with the DB2+ extractor turned
on, it will be ignored. In the case of DDF measurement, each SQL call
is intercepted for the requested duration, no sampling takes place.
The number of samples will always be converted to approximately one
per second.
- Measure to step end
- This field is not applicable when measuring specific DB2 stored
procedures or user-defined functions.
Specify 'Y' in this field
to indicate that the measurement is to continue to the end of the
step even if the specified number of samples has been recorded. Measurement
will continue at the sampling rate calculated based on the specified
duration and number of samples.
- Duration (min:sec)
- Specify the duration of the measurement. You can specify the
value in seconds or in minutes and seconds. To specify the duration
in minutes and seconds, separate the minutes value from the seconds
value using a colon. If configured during installation, a default
value will be displayed in this field for NEW and TNEW requests. An
installation default value is used if you do not supply input in this
field. Examples:
- 135 specifies 135 seconds
- 2:15 specifies 2 minutes and 15 seconds
- 2: specifies 2 minutes
The measurement will proceed for the specified time and the
sampling rate will be established at a frequency that would perform
the full number of samples for the specified duration.
The measurement
will terminate before the duration ends if the job step ends first.
- Delay by (secs)
- This field is not applicable when measuring specific DB2 stored
procedures or user-defined functions.
Specify a delay time, in
seconds, to occur before initiation of the measurement. The delay
will occur starting at the time execution of the job step begins.
This cannot be specified for measurement of a job that is currently
active.
- Notify TSO User
- Specify a TSO user ID to be notified upon completion of the
measurement. Enter blanks in this field for no completion notification.
- Retain file for (days)
- Specify the number of days after completion of the measurement
for which the measurement file is to be retained. The file and all
information about the measurement will be deleted after this period.
Enter blanks or zero in this field for no automatic deletion of the
measurement data.
- USS observations
- Specify the maximum number of spawned address spaces or substeps
to measure for a USS observation, up to the maximum defined in the
system configuration. The same sampling frequency will be used for
each spawned address space or substep. Sampling overhead can be high
if several spawned address spaces are running simultaneously.
When
this field is specified, the collection of measurements will be grouped
under a USS master record on the Observation List panel.
Note:
When you are deciding what values to enter
in the Number of Samples and Duration fields, consider that Application Performance Analyzer does
not have unlimited resources to store and report measurement data.
Data spaces are used for collecting and reporting data. Extremely
large amounts of measurement data can cause Application Performance Analyzer to fail in
either the data collection or reporting process. Data space requirements
for measurement data vary widely depending on the type of job or region
being measured, data extractors selected, etc.
[ Top of Page | Previous Page | Next Page | Contents | Index ]