Widnows Scheduler Job keeps running even after DM jobs complete

[dgsininet]

We recently upgraded from Decision Stream to Data Manager 10.2 and it's been a hell!

Before I seek help with some of the issues we are facing, here are some details about the env.
1.   Data Manager 64 Bit 10.2 on Windows 2008 R2 64 Bit
2.   Staging and Data Warehouse: SQL 2012 Enterprise
3.   Source Databases: DB2 on Unix, SQL Server 2012 on Windows 2008, and Oracle 11g Peoplesoft hosted with Oracle
4.   ODBC Config: Using the ODBC administrator from within Data Manager (so as to use the right version of the ODBC admin)
5.   SQL 2012 64 bit client and drivers
6.   Oracle 11g 32-bit drivers (64 bit didn't work even after numerous attempts – eventually found a knowledge base article stating 64 bti DB2 on Unix platforms is not supported by DM, or maybe that was DM on unix platforms, anyway couldn't get 64 bit to work)
7.   DB2 32 bit Client and drivers (64 bit didn't work)
8.   Jobs are executed in a 32 bit mode in the execute dialog as a 64 bit execution mode won't work with 32 bit ODBC drivers. So much for a 64 bit Data Manager install!! Really pi$$ed with this as the 32 bit execution mode also means that process memory is capped limiting our data load volumes.



Amidst many other issues, the one that I need help immediately with is below:

A scheduled job in Windows scheduler triggers the 32 bit rundsjobs.exe and calls the parent jobstream. This parent jobstream then fans out! Problem is, even after the jobstream completes, the scheduler job continues to show "running" and there's a process rundsjob.exe that keeps running. I checked Audit on parent jobstream and it shows everything green and complete with nothing running. I have checked the databases as well and they do not show any processes running from Data Manager against them. If I kill the rundsjobs.exe, then the scheduler stops and switches to "ready" mode.  For some reason, the rundsjobs.exe does not stop executing even when all the jobs are complete.

Any ideas what could be going on?

Thanks,
D

[pbose05]

I am also facing the same issue.

Can anyone kindly help on this.

[MFGF]

I am also facing the same issue.

Can anyone kindly help on this.

I have a sneaky feeling I know why this is happening. Can you take a look at the command you used in the scheduler to fire off the jobstream? Specifically, does it still have a -P argument specified?

eg

"c:\program files\ibm\cognos\c10_64_dm\bin\rundsjob.exe" -P "OLEDB" "User Id=sa;Password=~0085FD46CB00236AD837B4B0C8519CF917;Database=DMCat540;Data Source=(local);Provider=sqlncli11 (SQL Server 2012 native client);Integrated Security=FALSE" "Sound"

If so, this is your problem. The -P switch initiates a pause at the end of the jobstream. If you are running from a command window, you see a "Press Enter to continue" message, but obviously if this is being scheduled it will just sit there waiting for ever because there's nobody there to press Enter.

The simple solution is to remove the -P from the command :)

MF.

[pbose05]

Thank you MFGF...i missed that

Its working fine now.