Chapter 29: Monitoring and Reporting on Workflow
Monitoring Tools
Monitoring workflow helps you to identify problems with the current or predicted
workflow, so that you can resolve those problems. You can use the following CA
Workload Automation AE tools to monitor workflow:
Forecast Reports
Generate reports that display information about the predicted workflow to identify
problems before they occur.
Note: You can also use forecast reports to plan changes to your workflow in a test
environment.
Monitors
Track events to identify problems as they occur.
Browsers
Generate reports that display information about past events to identify recurring
problems.
You can solve problems before they occur or as they occur when you can identify the
issue that is associated with the problem. When you cannot determine the cause of a
problem, notify the administrator.
To solve a problem that you identify in real time using a monitor, correct the associated
issue and restart the job. To address recurring problems or problems with predicted
workflow that you identify using browsers and forecast reports, correct the associated
issues and use monitors to track the progress of the workflow.
Correcting issues that cause jobs to fail requires modifying workflow objects (job
definitions, machine definitions, and calendar object definitions). You can modify
workflow objects only when you have write access to those objects. When you cannot
solve a problem without modifying a workflow object and you do not have write access
to the problematic object, notify the scheduler.
Important! Modifying workflow objects sometimes has unexpected impacts on the rest
of the workflow. We recommend that you plan changes to the workflow in a test
Run a Monitor or Browser
Notes:
■ You can also use the CA WCC Forecast and Monitoring applications to monitor
workflow. For more information about these applications, see the CA WCC
documentation.
■ You can use the sendevent command to restart a job. You can use jil to update job
or machine definitions and the autocal_asc command to update calendar
definitions. For more information about the sendevent, jil, and autocal_asc
commands, see the Reference Guide.
Run a Monitor or Browser
Monitors and browsers help you identify problems with your workflow by tracking
events in real-time (monitors) or generating reports that display historical information
about events (browsers). For example, the scheduler component automatically restarts
jobs that fail, so examining events that change the status of jobs to RESTART helps
identify problems with specific jobs.
Note: By default, CA Workload Automation AE is configured to specify the date and time
using the "MM/dd/yyyy HH:mm[:ss]" format, but you can configure CA Workload
Automation AE to use a different format by changing the value of the DateFormat
parameter in the configuration file. For more information about the parameters in the
configuration file, see the Administration Guide.
Limits:
■ MM: 01-12
■ dd: 01-31
■ yyyy: 1900-current year
■ HH: 00-23
■ mm: 00-59
■ ss: 00-59
Follow these steps:
1. Open the operating system or instance command prompt:
■ (UNIX) Run the shell that is sourced to use CA Workload Automation AE.
The UNIX operating system command prompt opens. The shell that is sourced
to use CA Workload Automation AE presets all of the environment variables for
the instance.
■ (Windows) Click Start, Programs, CA, Workload Automation AE, Command
Prompt (instance_name).
The CA Workload Automation AE instance command prompt opens. The
command prompt presets all of the environment variables for the instance.
Run a Monitor or Browser
2. Enter the following command:
jil
The JIL command prompt for the local CA Workload Automation AE instance opens.
3. Define the monitor or browser by entering the following jil command and
attributes:
insert_monbro: monbro_name
mode: monitor | browser
[alarm: y|n]
[status_keyword: y]
[status_keyword: y]
[...]
all_status: y|n
all_events: y|n
currun|after_time: date_time
alarm_verif: y|n
Notes:
■ The utility also accepts values of m (for monitor) and b (for browser) in the
mode attribute of the jil monbro definition.
■ For more information about the insert_monbro subcommand and the jil
attributes that you can specify when you define a monitor or browser, see the
Reference Guide.
status_keyword: y
(Optional) Specifies whether or not to track status change events that are
associated with a particular job state. You can set multiple individual status
filters to track status change events that are associated with any combination
of the following job states:
■ RUNNING
■ SUCCESS
■ FAILURE
■ TERMINATED
■ STARTING
■ RESTART
The status keyword is the name of the job state in lower-case letters. To track
status change events that are associated with one of these job states, enter the
corresponding status keyword as the jil attribute and set the value of that
attribute to y.
Example: To track status change events that are associated with either the
FAILURE state or the TERMINATED state, specify the following attributes and
values:
failure: y
terminated: y
all_status: y|n
Specifies whether to track or report on all status change events. To track or
report on only status change events that are associated with certain job states,
set the value of the attribute to n and specify other status filters. Setting the
value of this attribute to y overrides other status filters.
all_events: y|n
Specifies whether to track or report on all events. To track or report on only
certain types of events, set the value of the attribute to n and specify other
event filters. Setting the value of this attribute to y overrides other event
filters.
currun: y|after_time: date_time
(Browsers) Specifies the historical time frame on which the browser reports. To
display only events that occurred during the most recent job run, set the value
of the currun attribute to y. To display all events that occurred after a particular
date and time, specify the after_time attribute. To specify the after_time
attribute, ensure that the currun attribute is set to the default value of n.
Note: By default, CA Workload Automation AE deletes archived event data that
is older than seven days, but you can configure the product to store data for a
longer period of time by editing the DBMaint script. The location of this script is
specified in the DBMaintCmd parameter of the configuration file. Specify a
value for the after_date attribute that is within the past seven days or edit the
script. For more information about the DBMaintCmd configuration parameter
and the DBMaint script, see the Administration Guide.
Example: To generate a report that displays a list of all processed events that
occurred after January 2, 2012 at 5 p.m., specify the following jil attributes:
currun: n
after_time: "01/02/2012 17:00:00" Run a Monitor or Browser
alarm_verif: y|n
(Monitors) Specifies whether the monitor waits for you to acknowledge alarms
that the scheduler issues. This attribute is valid only for monitors that are
defined to track alarms. When you set the value of this attribute to y, you are
prompted to acknowledge any alarms that the scheduler sends.
A monitor is defined to track alarms when the value of the alarms attribute is
set to y or when the value of the all_events attribute is set to y. For monitors
that are defined to prompt you to acknowledge alarms, the monbro utility
pauses, displays the name of the alarm, and issues the following message and
prompts:
Alarm: alarm_name issue MM/dd/yyyy HH:mm:ss Run# run_num
Message Acknowledged by: user_name
Comment:[your_comment]
The utility displays the alarm information and leaves the following two fields
blank. To acknowledge the alarm, enter your user name in the Message
Acknowledged by field. Optionally, you can enter information in the Comment
field (such as "Reported to DB Administrator for follow up").
Important! The monitor cannot resume until you acknowledge the alarm by
entering your user name.
Note: For more information about the monbro utility, see the Reference Guide.
4. Enter the following jil command:
exit
The monitor or browser is inserted into the database, and the jil command prompt
closes.
5. Open the operating system or instance command prompt and enter the following
command:
{monbro -N monbro_name|monbro -N monbro_name -P ss}
-P ss
(Optional, Monitors only) Specifies the frequency, in seconds, at which the
monitor polls for events. The monitor polls for events at the default frequency
when you do not specify this option. The option is not valid with browsers.
Limits: Integers greater than 0
Default: 10
The monitor or browser runs. A monitor polls for events at the specified interval. A
browser generates a report that displays the event information for the specified
period based on the specified event filters.
Example: Define a Browser
This example defines the browser named job_restarts to generate a report listing all
events that changed the status of a job to RESTART after September 10, 2012 at
midnight.
insert_monbro: job_restarts
mode: browser
restart: y
all_status:n
all_events:n
after_time: "09/10/2012 00:00:00"
Example: Define a Monitor
This example defines the monitor named track_alarms to track alarm events and to
prompt the user to acknowledge all alarms as they occur.
insert_monbro: track_alarms
mode: monitor
alarm: y
all_status: n
all_events: n
alarm_verif: y
Example: Acknowledge an Alarm
This example shows that the operator acknowledged notification of a JOBFAILURE alarm
when the job with run number 33:2 failed at 20:15 29 on 09/24/2012 and that the
operator reported the job failure to the scheduler for follow up and resolution.
Alarm: JOBFAILURE fail 09/20/2012 20:15:29 Run# 33:2
Message Acknowledged by: operator
Comment:Reported to the scheduler for follow up.
Example: Run a Monitor
This example runs the monitor named track_alarms so that it polls the database for
alarm events at 30 second intervals.
monbro -N track_alarms -P 30 Generate a Forecast Report
Generate a Forecast Report
Forecast reports display information about predicted workflow. Forecast reports help
you identify problems with the predicted workflow to resolve them before they occur or
to plan changes in the workflow.
Note: By default, CA Workload Automation AE is configured to specify the date and time
using the "MM/dd/yyyy HH:mm[:ss]" format, but you can configure CA Workload
Automation AE to use a different format by changing the value of the DateFormat
parameter in the configuration file. For more information about the parameters in the
configuration file, see the Administration Guide.
Limits:
■ MM: 01-12
■ dd: 01-31
■ yyyy: 1900-current year
■ HH: 00-23
■ mm: 00-59
■ ss: 00-59
Follow these steps:
1. Open the operating system or instance command prompt:
■ (UNIX) Run the shell that is sourced to use CA Workload Automation AE.
The UNIX operating system command prompt opens. The shell that is sourced
to use CA Workload Automation AE presets all of the environment variables for
the instance.
■ (Windows) Click Start, Programs, CA, Workload Automation AE, Command
Prompt (instance_name).
The CA Workload Automation AE instance command prompt opens. The
command prompt presets all the environment variables for the instance.
2. Enter the following command:
forecast {-M machine_name| -J job_name [-M machine_name]} -F "mm/dd/yyyy HH:MM
[:ss][-T "mm/dd/yyyy HH:MM [:ss]]" [OPTIONS]
-M machine_name| -J job_name [-M machine_name]
Specifies what predicted workflow information the forecast report displays. To
specify predicted workflow for a particular machine, use the -M parameter. To
specify predicted workflow for a particular job, use the -J parameter.
Notes:
■ The report displays predicted workflow for all jobs that are scheduled to
run on the specified machine when you specify the -M parameter alone.
■ The report displays a predicted workflow that lists runs of the specified job
on all machines when you specify the -J parameter alone.
■ The report displays a predicted workflow that lists runs of the specified job
on the specified machine and excludes runs on other machines when you
specify the -J and -M parameters together.
-F "MM/dd/yyyy HH:mm [:ss]"
Specifies that the forecast report predicts workflow starting on the specified
date.
-T "MM/dd/yyyy HH:mm [:ss]"
(Optional) Specifies that the forecast report predicts workflow ending on the
specified date. This option is required only when you want to forecast
workflow for multiple days.
OPTIONS
(Optional) Specifies optional parameters that you can use to control what
information the forecast report displays.
Note: For more information about these optional parameters, see the
Reference Guide.
The command generates a forecast report that displays information about
predicted workflow.
3. (Optional) Depending on the information that you need, repeat the procedure for
other jobs or machines that are defined within the instance.
The command generates additional forecast reports.
No comments:
Post a Comment