Focal Point
Next Run Time When Using &&KILL_RPC

This topic can be found at:
https://forums.informationbuilders.com/eve/forums/a/tpc/f/7971057331/m/5291076603

November 12, 2008, 11:24 AM
jmdoyl1
Next Run Time When Using &&KILL_RPC
I was reading the New Feature Document for 7.6.7 and for report caster there is an option to set the &&KILL_RPC_NRT. I wanted to get the groups input on if this would be a way to address a need we have.

Self Service Application (JSP, uses caster api), 7.1.3 Windows Web Unix Report Server/Distro Server.

We have ETL jobs that run each night. The scheduled reports (defined by the end user/admin to run daily, weekly, monthly etc.) The jobs are scheduled to run when batch should be done. However, some times it runs late and we shut down the distro server so that jobs that would have been que'd up to run don't execute.

Now that the &&KILL_RPC_NRT can be set my initial thought was to add a pre-processing procedure to the scheduled jobs to check if batch is finished. If not then we can kill the job but leave the nextrun time alone so that it will wake up and try again, and again and again (one draw back that I see a lot of overhead trying to run while waiting on batch). But we would not have to shut down distro and other jobs that are not dependant on batch or new jobs that need to be scheduled would not be affected.

Thanks in advanace for your input.


WebFOCUS 7.1.3 Windows Client AIX 5.2 Report Servers
November 16, 2008, 02:21 PM
FortuneCookie
I'm all in favor of automation. It beats having to remember whether or not you've shut down the distribution server.

It's kind of like leaving the house wondering if the iron was on/off. Eventually you'd have to get out of your way and check.