IBM Support

QMGTOOLS: Job Watcher Monitor Function (Continuous Job Watcher)

Question & Answer


Question

Is there a continuous Job Watcher function?

Answer

QMGTOOLS provides a program that allows for a continuous Job Watcher (JW) collection that collects and keeps a certain amount of collections.

Installing QMGTOOLS
If QMGTOOLS is not installed, follow this URL :

http://www-01.ibm.com/support/docview.wss?uid=nas8N1011297

Usage

Step 1 - Go to the Performance/Misc menu

To get to the Performance menu, type command GO QMGTOOLS/MG and the main Must Gather menu should be available like the example below. In this example, the Performance/Misc menu is option 3.


Step 2 - Select the Monitors section and then select Continuous JW trace


Performance menu

Continuous JW trace menu


Step 3 - Parameters

After selecting the Continuous JW trace, there is two options. *START to start the trace or *STOP to stop the trace.

When selecting *START, several parameters will appear in the next screen.



 
Size of Job Watcher per collection The size of the collection in gigabytes
Second per collection How many seconds to run Job Watcher cycles to a new collection. Default is 1800 seconds or 30 minutes
Collection interval Job Watcher's main function is to take a snapshot of the system every xx seconds. Default is 10 seconds.
JW collection to keep How many collections to keep. Default is 2.
Library for collection The library to store the JW data (collection). Default IBMJW.

Note: Job Watcher will write to this library, so if any save is performed, omit this library if this program is active.
Collect SQL statements Collect SQL statements.
*NONE - does not collect any SQL statements
*LOW - collects *SQLCURSTMT *ALWAYS
*MED - collects *SQLSTMT *ALWAYS
*HIGH - collects *SQLDETAIL *ALWAYS
Include inactive jobs/tasks *YES, *NO

Specifies whether detailed data for all jobs and tasks, including inactive jobs and tasks, will be collected for
the first interval.
Dump Plan Cache end of each interval Y, N

After the end of each JW collection (as defined in seconds per collection parameter), submit a job to dump the plan cache. Default is N.
# of Plan Cache to keep in /tmp/sqe How many plan cache dumps to keep. The plan cache is stored in the IFS under /tmp/sqe
F-keys
F1 = Continue Continue and run the continuous JW program
F6 = Save and Exit To change defaults to parameters listed here, press F6 to save and exit the program. You will need to start again via *START command if you want to run the program.

You would normally do this if you would like to start the continuous JW trace without being prompted by this parameter screen.


Step 4 - Starting the job

From Step 3, after pressing F1, the continuous Job Watcher will be submitted. The jobs that are submitted will be:

JWMON jobxxxx BATCH ACTIVE PGM-QJWMON
IBMJW jobxxxx BATCH ACTIVE PGM-QJW

Note: If you want to start the continuous JW without going through the parameter screen, then submit the command the example below. However, before doing so, please verify the parameters from Step 3 and hit F6 to save/exit so the new defaults can be saved.

SBMJOB CMD(QMGTOOLS/QJWMONCMD FUNCTION(*START)) JOB(QJWMON) JOBQ(QSYSNOMAX)


Step 5 - Ending the JW continuous job

QMGTOOLS/QJWMONCMD FUNCTION(*STOP)

 
Where the data is store

JW location

The JW data is located in the library specified by the parameter Library for collection. You can save that library.
The preferred data upload method is QMGTOOLS: FTP Data to IBM
http://www-01.ibm.com/support/docview.wss?uid=nas8N1010370

Plan Cache location

If using Dump Plan Cache option, that data will be stored in the IFS directory under /tmp/sqe. There should be some sub directory or directories (depending on how many plan cache dumps to keep) with a naming convention starting with jobuser_jobnumber_date_time. The jobuser is the user that started the JW continuous trace. To view that directory run command WRKLNK '/tmp/sqe' .

In the sample below, directory /tmp/sqe/SYSADMIN126647_12302014095448 is the plan cache that was collected from the JW continuous trace. To get the full path, put an option 8 (display attributes) next to the directory.



To save the objects, create a save file and use the SAV command like the example below:

SAV DEV('/qsys.lib/savelib.lib/savefile.file')
OBJ(('/tmp/sqe/SYSADMIN126647_12302014095448'))

where savelib is the location of the save file and savefile is the name of the save file.

[{"Product":{"code":"SGYQGH","label":"IBM i"},"Business Unit":{"code":"BU009","label":"Systems - Server"},"Component":"Performance","Platform":[{"code":"PF012","label":"IBM i"}],"Version":"Version Independent","Edition":"Standard"}]

Document Information

Modified date:
30 July 2019

UID

nas8N1020478