Defaulting to a server group while scheduling

All,

We have a BOXIR2 Enterprise edition with clustered server. While scheduling a report, We have given different groups of users the right to select different server groups so that their reports get scheduled via only their particular assigned server group. But for this, while scheduling user needs to select " Use the servers belonging to selected group option" and then choose their server. If they do not choose this option the report will by default run on first available server.

Is there a way that even without the user selecting this “select server group” option , the report is automatically run only on the assigned server group?

Any help would be greatly appreciated.

Thanks,
M


marym (BOB member since 2006-09-21)

You can set the Default Server Group for a report - in the CMC go into the properties of a report, on the “Process” tab, you can set the Server Group which will be used when the report is scheduled (Although users can still change this to the wrong Server Group at schedule time).


blair :canada: (BOB member since 2006-11-17)

MaryM and Blair, we too would like to do this but I don’t want to have to touch all 1500+ scheduled reports in the CMC.

We have a clustered environment with 4 severs and I want to completely separate all report processing in XIr3.1 onto 1 or 2 servers and NOT have any scheduled reports running on the other two boxes.

Do you know of a way I can simply create a single server group that uses the two “scheduled report” servers I’ve set up and force all reports to schedule there?

Thanks and regards,
Don


don.davis :us: (BOB member since 2006-09-06)

When you designed your cluster, did you install and enable the various JOB servers on all of the physical servers…?

Look at the process-flows for the various JOB types and only install / enable the JOB servers on the two (2) servers where you want them to RUN.

In the BOBJ Admin docs this is referred to as the “processing layer”.


MJRBIM :canada: (BOB member since 2007-03-23)

MJRBIM, in xir2 that’s exactly how I would have done it. in r3 the WebIntelligenceProcessingServer is needed for both ad-hoc and scheduled reports, or that’s my understanding. So I can lump ALL report processing together on segragated servers in the cluster, but I’d rather not.

If i’m wrong about the processing, please correct me.

Don


don.davis :us: (BOB member since 2006-09-06)

You are correct, the WebIntelligenceProcessingServer is used for SCHEDULED processing as well.

You didn’t mention WebI, and I assumed CRYSTAL for the reports.

So what you are trying to do is have two (2) separate WebI processing paths - one for On-Demand/Ad-Hoc, and one for SCHEDULED instances, right…?

What sort of volume are you expecting for both paths - and how many servers do you have in the cluster…?


MJRBIM :canada: (BOB member since 2007-03-23)

we’ve got about 1600 recurring reports, many of which are daily. user base is 4k+, with about a max of 300 concurrent active at peak. 4 boxes in the cluster, and we wish to use two of them for scheduled report activity in order to completely segreg te it, due to previous performance issues in r2. Does that help?

Don


don.davis :us: (BOB member since 2006-09-06)

The only way I can think of doing it is creating your server group, then RESCHEDULING you 1,600 report instances to assign them to the new server group.

Are users able to schedule their own reports…? If YES, then you will also need to set the DEFAULT server group on each object.

Considering the size of your deployment, you might want to write an SDK application to do these tasks - or go with something like the APOS tools…

Object Manager
Instance Manager
Info Scheduler


MJRBIM :canada: (BOB member since 2007-03-23)

was really hoping to avoid touching all those reports, for obvious reasons. yes, users are free to schedule, which is why I really want to remove as much guesswork (read: “choice”) as possible for everyone. Thanks!

Don


don.davis :us: (BOB member since 2006-09-06)