Is there a way to setup RC to automtically restart a job that aborts? I have 2 conditions:
1. The first of the month the Oracle Database is taken down. So jobs scheduled during that time abend.

Yea I could not schedule any job to run during that time but I'm afraid I'll run out of time slots, plus some jobs need to be run by 6 AM for example. So the question is can RC try to rerun the job later?

Message Code Message Text

BTP1010 Schedule Executed Due To NEXTRUNTIME
BTP1020 Starting task: TS0012DB - Daily Feed Report
BTP1020 Task type: MR Standard Report
BTP1020 Task domain: business/business.htm
BTP1020 Retrieving MR report: app/ts0012db
BTP1020 Connecting to server WFNT with execution id MANUSZR
BTP1020 Executing focexec.
BTP1020 TBRACCD_FEED_DATE = 2012/05/01
BTP1020 SQLCODE IS 1034 (HEX: 0000040A)
BTP1020 CONNECT FAILURE
BTP1020 : ORA-01034: ORACLE not available
BTP1020 : ORA-27101: shared memory realm does not exist
BTP1020 : SVR4 Error: 2: No such file or directory
BTP1020 L (FOC1406) SQL OPEN CURSOR ERROR. : TBRACCD
BTP1010 Task error: CONNECT FAILURE
BTP1010 No report to distribute.

2. Report Caster goes down.
So when it get restarted can jobs that were suppose to run during that time be automatically started. Kind of a look back issue.
Thanks
Rick

This message has been edited. Last edited by: Rick Man,

do you know about 'pre processing conditions'?
Tasks/Edit Tasks/Advanced/Procedures....
you can write up to 2 procedures that run before the caster job itself. [nb: put them in the baseapp.]. This job/jobs could check some status code somewhere, and if you got a good result, then go ahead and run the caster job, but if you get a bad result, then this little fex says:

-SET &&KILL_RPC =  'Y';

and then the caster job itself dies, and doesn't run that time.
so....
you could set your job to run alot...often...sevaral times a day... and its pre-processing fex could check , say, the UPDATE_DATE_TIME in some Oracle file. And if the UPDATE_DATE_TIME is current, then set the KILL_RPC to Y. If that date is old, or yesterday, or whatever you decide, then SET &&KILL_RPC to N or jsut comment out the line.
..

This way, you get mutliple chances during a day to run your job. If all is well, only 1 will execute. If all is not well, then you'll get many more chances to run.
...
Make sure you research and understand the SCANBACK feature of Caster, not for this question, but just so you don't get confused.
..
Howzat?

Susannah...

If it is in the current day then the Report Caster will look for all schedules that have a NEXRUNTIME <= SYSTEM TIME.

So, If your job is scheduled at 5 am and if you have disabled your scheduled before 5 am and enabled it at 6 am and it will automatically run the job. Same as the case even if you stop RC and re-start.

Note: I am not saying about the check box to enable in the schedule. i am taking about using the ACTIVE field in the BOTSCHED.

When possible better to use the pre-processor option as Susan mentioned.

Meet us at Summit 2012 : Casestudies: "Customizing ReportCaster"

Thanks for the input. Rao, If the job abends the nextruntime is set for the next run, so restarting the server or doing a scanback doesn't work. I will look into recovery and scanback more maybe it will at least allow me to run all jobs from point A to point B and not have to resubmit them one at a time.