| October 16, 2020
Creating a File Dependency for a Job Stream using Start Condition
Waiting for a file to arrive before a job can start processing it is the most common and quintessential requirement for any workload automation tool. Historically, this was done by creating a File dependency using OPENS and Unix tricks to manage wild cards and multiple matching files. Then Event Driven Workload Automation was introduced where an Event Rule could monitor a file with wild cards and when the condition was satisfied the dependent Job Stream was submitted. With the advent of the Start Condition feature introduced in version 9.4 Fix Pack 1, the event rule file monitor capability is integrated into a Job Stream using the utility, filemonitor. Another important aspect of this requirement is to pass the name of the matching file to another job in the workflow to process the data contained in the file.
This article explores this feature and also uses an internal variable, ListOfFilesFound, to pass the name of the matching file to a successive job in the same Job Stream.
How to use the Start Condition feature with File Created
Follow the steps below to create a Start Condition for a Job Stream
- Login to the Dynamic Workload Console
- Navigate to Design à Manage Workload Definitions
- Select Create New à Job Stream
- In the General tab, fill in the Job Stream name and Workstation
- In the Start condition tab, fill in the values in the fields as shown below:
- Condition: File created An existing file with a matching name will not satisfy the condition, a new file with a matching name must be created.
- Workstation: EEL
- File: /tmp/start.cond.*.txt
- Output file: /tmp/start.cond.out.txt
- Job name: FILE_CREATED A file monitor job with this name is auto generated. This job iteratively keeps monitoring the file and when satisfied, submits a new instance of the Job Stream.
6. Select Create New à Job Definition à UNIX
7. In the General tab, fill in the Job name, Workstation, and Login
8. In the Task tab, enter the following
Script name: /bin/echo “File Monitored: ${job:FILE_CREATED.ListOfFilesFound}”
9. Click on Save
10. Select Create New à Job Definition à UNIX
11. In the General tab, fill in the Job name, Workstation, and Login
12. In the Task tab, enter the following
Script name: cat /tmp/start.cond.out.txt
13. Click on Save
14. Add the two jobs created earlier to the Job Stream
15. Click on Save
16. The following is the Job Stream definition is composer format.
| SCHEDULE EEL#START_COND_JS STARTCOND FILECREATED EEL#”/tmp/start.cond.*.txt” INTERVAL 60 ( ALIAS FILE_CREATED RERUN OUTFILE “/tmp/start.cond.out.txt” ) : EEL#FILE_NAME_FROM_VARIABLE SCRIPTNAME “/bin/echo \”File Monitored: ${job:FILE_CREATED.ListOfFilesFound}\”” STREAMLOGON iwadmin DESCRIPTION “Retrieve file name in Start Cond from an internal variable, ListOfFilesFound” TASKTYPE UNIX RECOVERY STOP
EEL#FILE_NAME_FROM_OUTPUT_FILE SCRIPTNAME “cat /tmp/start.cond.out.txt” STREAMLOGON iwadmin DESCRIPTION “Retrieve file name in Start Cond from the Output file field” TASKTYPE UNIX RECOVERY STOP FOLLOWS FILE_NAME_FROM_VARIABLE
END |
17. Select the Job Stream Name and choose Select an Action à Submit Job Stream into Current Plan
Monitor the Status of Job Stream with the Start Condition and File Created
Follow the steps below to monitor the status of the Job Stream
- Login to the Dynamic Workload Console
- Navigate to Monitoring and Reporting à Monitor Workload
- Choose Job Stream for Object Type and click on Run
- In the Filter window, enter Start_cond and press Enter
- In the resulting view, Right Click on the Job Stream, START_COND_JS, and choose Jobs
- Notice that there are 4 jobs in the Job Stream
FILE_CREATED Auto generated job to keep monitoring the file
RESTART_STARTCOND Auto generated job to submit the Job Stream
FILE_NAME_FROM_VARIABLE Job to retrieve the matching file from a variable
FILE_NAME_FROM_OUTPUT_FILE Job to retrieve the matching file from an output file
7. Left click on the Job, FILE_CREATED, to view its properties. Click on Extra Properties tab to view the internal variable that stores the name of the matching file.
8. Right click on the Job, FILE_NAME_FROM_VARIABLE, and choose Job Log. Notice that the name of the matching file is retrieved from the internal variable, ListOfFilesFound, set by the job,
FILE_CREATED, ${job:FILE_CREATED.ListOfFilesFound}.
File Monitored: /tmp/start.cond.092920201352.txt
9. Right click on the Job, FILE_NAME_FROM_OUTPUT_FILE, and choose Job Log. Notice that the name of the matching file is retrieved from the file name specified in the Output file field,
cat /tmp/start.cond.out.txt
/tmp/start.cond.092920201352.txt
How to use the Start Condition feature with File Modified
Follow the steps below to create a Start Condition for a Job Stream
- Login to the Dynamic Workload Console
- Navigate to Design à Manage Workload Definitions
- Select Create New à Job Stream
- In the General tab, fill in the Job Stream name and Workstation
- In the Start condition tab, fill in the values in the fields as shown below:
- Condition: File modified An existing file with a matching name will not satisfy the condition, the file with a matching name must be modified.
- Workstation: EEL
- File: /tmp/start.cond.*.txt
- Output file: /tmp/startcond.out Must have a different pattern than the file being monitored.
- Additional Parameters: -modificationCompletedTime 120 When a file is modified, the event is not sent immediately, but only after the interval of time specified by -modificationCompletedTime <seconds> has elapsed and during which no additional changes were made to the file. For other parameters, refer to Chapter 16, Using utility commands, in the IBM Workload Scheduler version 9.5 User’s Guide and Reference.
- Job name: FILE_MODIFIED A file monitor job with this name is auto generated. This job iteratively keeps monitoring the file and when satisfied, submits a new instance of the Job Stream
6. Select Create New à Job Definition à UNIX
7. In the General tab, fill in the Job name, Workstation, and Login
8. In the Task tab, enter the following
Script name: /bin/echo “File Monitored: ${job:FILE_MODIFIED.ListOfFilesFound}”
9. Click on Save
10. Select Create New à Job Definition à UNIX
11. In the General tab, fill in the Job name, Workstation, and Login
12. In the Task tab, enter the following
Script name: cat /tmp/startcond.out
13. Click on Save
14. Add the two jobs created earlier to the Job Stream
15. Click on Save
16. The following is the Job Stream definition is composer format.
| SCHEDULE EEL#START_COND_JS STARTCOND FILEMODIFIED EEL#”/tmp/start.cond.*.txt” INTERVAL 60 ( ALIAS FILE_MODIFIED RERUN PARAMS “-modificationCompletedTime 120” OUTFILE “/tmp/startcond.out” ) : EEL#FILE_NAME_FROM_VARIABLE SCRIPTNAME “/bin/echo \”File Monitored: ${job:FILE_MODIFIED.ListOfFilesFound}\”” STREAMLOGON iwadmin DESCRIPTION “Retrieve file name in Start Cond from an internal variable, ListOfFilesFound” TASKTYPE UNIX RECOVERY STOP
EEL#FILE_NAME_FROM_OUTPUT_FILE SCRIPTNAME “cat /tmp/startcond.out” STREAMLOGON iwadmin DESCRIPTION “Retrieve file name in Start Cond from the Output file field” TASKTYPE UNIX RECOVERY STOP FOLLOWS FILE_NAME_FROM_VARIABLE
END |
17. Select the Job Stream Name and choose Select an Action à Submit Job Stream into Current Plan
Monitor the Status of Job Stream with the Start Condition with File Modified
Follow the steps below to monitor the status of the Job Stream
- Login to the Dynamic Workload Console
- Navigate to Monitoring and Reporting à Monitor Workload
- Choose Job Stream for Object Type and click on Run
- In the Filter window, enter Start_cond and press Enter
- In the resulting view, Right Click on the Job Stream, START_COND_JS, and choose Jobs
- Notice that there are 4 jobs in the Job Stream
FILE_MODIFIED Auto generated job to keep monitoring the file
RESTART_STARTCOND Auto generated job to submit the Job Stream
FILE_NAME_FROM_VARIABLE Job to retrieve the matching file from a variable
FILE_NAME_FROM_OUTPUT_FILE Job to retrieve the matching file from an output file
7. Left click on the Job, FILE_CREATED, to view its properties. Click on Extra Properties tab to view the internal variable that stores the name of the matching file.
8. Right click on the Job, FILE_NAME_FROM_VARIABLE, and choose Job Log. Notice that the name of the matching file is retrieved from the internal variable, ListOfFilesFound, set by the job,
FILE_CREATED, ${job:FILE_MODIFIED.ListOfFilesFound}.
File Monitored: /tmp/start.cond.092920201352.txt
9. Right click on the Job, FILE_NAME_FROM_OUTPUT_FILE, and choose Job Log. Notice that the name of the matching file is retrieved from the file name specified in the Output file field,
cat /tmp/start.cond.out.txt
/tmp/start.cond.092920201352.txt
Want to learn more? Write me at Sajjad.Kabir@hcl.com or schedule a demo.
Author’s BIO
Sajjad Kabir, Solutions Architect, HCLSoftware
Sajjad Kabir is an Information Technology professional with over 25 years of diverse industry experience with primary emphasis in IT Architecture and Service Management. Extensive experiences in management, architecture, design, development, implementation, and systems integration solutions in multiple industries, platforms, and network environments.
View my profile on LinkedIn
Comment wrap
