13.1.1 Acquiring information about job execution activity by specific users
When a company's information systems department uses JP1/AJS3 - Manager on an integrated basis, you might need to find out who uses JP1/AJS3 and to what extent by examining the history of job activity for individual users. This case study describes how to collect a history of job activity at the user level.
- Organization of this subsection
(1) Background
Traditionally, each department in an organization built and operated its own systems. However, this approach incurred significant operating costs and offered no benefits in terms of information sharing and resource allocation. To overcome this issue, organizations are now optimizing information systems through consolidation by making the applications in the information systems department available to all departments.
(2) Goals
With JP1/AJS3 being used on an integrated basis in the information systems department, you want to collect information that identifies who executed what job, when and where the job was executed, and how long it ran.
(3) Approach
By viewing the JP1 events (job normal end events, job abnormal end events, and job warning end events) output by JP1/AJS3, you can acquire information about the circumstances surrounding the execution of a particular job.
The following table lists the aspects of job history you can acquire, and the JP1 event attributes that correspond to each type of information.
No. |
Information sought |
Corresponding JP1 event attribute |
---|---|---|
1 |
Who? |
|
2 |
When? |
|
3 |
What? |
|
4 |
Where? |
|
5 |
How long? |
Job execution time at the execution destination# |
- #
-
Output by JP1/AJS3 - Manager version 09-50 or later.
(4) Function overview
Users can compile the history of job activity in any format by using the jevexport command to output the JP1 events output by JP1/AJS3 to a CSV file.
(5) Configuration example
This example shows how to configure JP1/AJS3 to collect the history of job activity for the week (from 7 days ago until the previous day) at 1:00 a.m. every Sunday, from August 1st 2011.
-
Create a filter file to be used with the jevexport command.
Configure the filter file to do the following:
-
Extract the following job end events:
- Job normal end events (event ID: 4106)
- Job abnormal end events (event ID: 4107)
- Job warning end events (event ID: 4109)
For details about the JP1 event attribute values, see A.2 JP1 event attributes.
-
Extract events generated from 7 days ago until the previous day (in effect the last 8 days excluding today).
For example, create the filter file as follows:
B.ID IN 4106 4107 4109 B.TIME WITHIN D 8 EXCLUDE B.TIME WITHIN D 1
For details about filter files, see the JP1/Base User's Guide.
-
-
Create a jobnet in JP1/AJS3.
Create a jobnet with the following characteristics:
- The schedule rules of the root jobnet
-
Set the following schedule:
- Start date: First Sunday in August, 2011
- Start time: 01:00
- Processing cycle: Weekly
- Detailed definition of the PC job
-
Define the job as follows:
- File name: jevexport
- Parameters: -o output-file-name -f filter-file-name
-
Register the jobnet you created in step 2 for planned execution.
The PC job will execute at 1:00 a.m. every Sunday after August 1, 2011. When the PC job finishes executing, the past week's history of job activity is output to the file specified by output-file-name.
-
Use a spreadsheet or other software to compile the information in the output file.
(6) Manual references
Type |
Item |
Location |
---|---|---|
Overview |
Filter file |
JP1/Base User's Guide |
JP1 event attribute values |
||
Commands (jevexport command) |
JP1/Base User's Guide |
(7) Cautionary notes
-
The JP1 event that reports the job execution time at the target host gets its value by subtracting the start time of the job from the end time. Therefore, the value of this attribute will be zero if the job runs for less than a second.
-
The JP1 event that reports the job execution time at the target host works in units of seconds. The effects of this rounding are cumulative, and the value becomes less accurate as the number of jobs increases. In this type of scenario, use the value as a reference only.