SLEEPER Documentation Help

[mandrew]

Hi Everyone,

I've been recently tasked with the job of documenting all of the jobs that are scheduled on our production iSeries server running JDE World 9.1. We've run into a serious case of "black box syndrome" which has resulted in our current SME's being left with an incomplete picture of how jobs interact and are scheduled.

I have little experience with JDE World, specifically the SLEEPER scheduling subsystem, and was wondering if there was a best practice when documenting jobs, their assigned queues , and any other details that may be important. A report that could be generated showing the scheduled times and dates/days of the week would be preferred.

Right now we have jobs running on both IBM's scheduler and SLEEPER that interact as well, so you can imagine the frustration we experience when a job fails and we need to spend a few hours manually stepping backwards through both schedulers to understand the nature of any problem that arises.

Any input would be greatly appreciated. As of now I've fallen back to using an excel spreadsheet to manually input and sort jobs across both schedulers.

Thanks in advance for your thoughts,

Andrew

 

[Eric Lehti]

Andrew,
You can document your SLEEPER scheduled jobs with this query (written either in World Writer or IBM Query)
Files to join
F01. F9501 Unattended Operations Scheduling Master
F02. F00821 Menu Selections LO Left Outer join
F03. F9801 Software Versions Repository Master

From Field File ID Relation To Field File ID
UNPID F01 EQ MZMNI F02
MZJTOE F02 EQ SIMID F03


Seq W P Col Sup Edt Prt Num Output
No. Description Size L W Sp Hdg Cde Dec Scl Field
10 Unattended Job Start Time 8 N A *DF N Z 0 0 UNJTTM
20 Program ID . . . . . . . 10 N A *DF N UNJPID
30 Program ID . . . . . . . 10 N A *DF N UNPID
40 Selection Number . . . . 5 N A *DF N 0 0 MZSELN
50 Job to Execute . . . . . 10 N A *DF N MZJTOE
60 Description. . . . . . . 60 N A *DF N SIMD

Here is sample output

Start Program Program Sel Job to Member
Time ID ID No Execute Description
------ ---------- ---------- ------ ---------- ------------------------------------
230600 J81900 %SLEEPER1 3 J55DLY05 Delay between jobs on the jobq
230600 J81900 %SLEEPER1 4 J553482 Master Planning Schedule
230600 J81900 %SLEEPER1 5 J55QSTR Start QINTER after customer invoicin
230600 J81900 %SLEEPER1 6 J42119TG F42119 tag-along 1-to-1
230600 J81900 %SLEEPER1 7 J42996 Purge Details to History
230600 J81900 %SLEEPER1 8 J4201P Sales Order Header (F4201) - Purge
230600 J81900 %SLEEPER1 9 J82001 Query Print Driver
230600 J81900 %SLEEPER1 10 J82001 Query Print Driver
230600 J81900 %SLEEPER1 11 J55OTOT update F4201 field SHOTOT and F42019
230800 J81900 %SLEEPER1Z 1 J55SLEEPER Allows you to run any Query/400 quer
230800 J81900 %SLEEPER1Z 2 J55SLEEPER Allows you to run any Query/400 quer
230800 J81900 %SLEEPER1Z 3 J55STR3 Update JDE from Star - Fax
230800 J81900 %SLEEPER1Z 4 J55STR4 Update JDE from Star - Fax
230800 J81900 %SLEEPER1Z 5 J55STR5 Update JDE from Star - PHONE





You want F9501 as the primary file because it is the SLEEPER file

The reason you join to F00821 with Left-Outer-join is to get capture any J81900 %percent% menu jobs (if your site uses %percent% menus in SLEEPER)
The reason you join to F9801 (Software version repository file) is to capture the program descriptions) This works IF your programs are developed and stored in SVR. IF NOT, you will need to DSPOBJD programs get the descriptions.

The way to integrate the IBM WRKJOBSCE jobs into this report would be to insert a "place holder job" record into SLEEPER F9501 that is a d0-nothing job other than to document that the IBM scheduler runs a job by that name at that time. A bit of double-data entry, but doable.

For additional descriptive text, you could join to a 4th file (Left Outer Join) that you keep additional descriptive text beyond what the object description provides.

 

[mandrew]

Thanks Eric! I'll pass this along to my team.

-Andrew