[b]Put all Scheduled reports on Hold[/b]

[roswald]

We have a reporting database that gets refreshed nightly. If for any reason the refresh didn't complete, crashed, etc. I would want to make sure no reports ran until the refresh was completed.
Can anyone tell me how this can be accomplished in CE9 or provide any workaround?


Thank you for helping!

 

[synapsevampire]

You can probably have the database stop the Crystal services on the server until after the job completes, I've done this before. If it fails, don't restart them, otherwise do.

You'll have to learn something about your database though, such as it's type and version.

-k

 

[ShortyA]

You could set the reports to have event-based triggers and then set the database to place a file on the CE server once it has refreshed. I have used this quite a bit for overnight datawarehouse refreshes and then kicking off a batch of reports.

ShortyA

 

[roswald]

ShortyA
Our Enterprise Reporting system has over 1000 reports that have to be taken into consideration.
Setting a trigger would be a good idea except there are days that the report runs but no backup, so then I would have to create 2 schedules for those. In the end, it becomes too much work.

Synapsevampire
I don't clearly understand your response. When you say, "have the database stop the Crystal services" are you talking about the reporting database or the server that CE9 is installed on? What exactly are Crystal services?

I do have a solution that I would like to share with you. If I stop the Crystal Job Server, no reports would run but users could still view report objects, view reports on demand, and schedule reports. The reports just won't run until after the Job Server is started. What do you think?

Thank you all for the suggestions.

B

 

[dhulbert]

roswold.

Thats what Synapsevampire was talking about, but dependin on th edatabsae you are using you might be able to set up a job that runs before and after the updates to stop and start the crystal server.

 

[mdwyer]

See Thread782-1037002 for ideas shared recently.

But I find file events difficult to manage with 100 reports. With your 1000, I would suggest you look at synapsevampire's idea, and your own.

 

[spedi]

The best thing is still the file based event driven. That's how we implement all our reports at run. Each database refresh and backup creates a file event and the reports connected to those events fire off. We have more than 200 reports and about 2000+ schedules per week. Hence I don't agree with the statement of MDWYER with respect to "find file events diffcult to manage with 100 reports"

No problem with Job Server concept; except if the end users have "on demand access" to the reports.

Srinath

http://www.nhsonline.org

srinath_p (at) yahoo.com

 

[mdwyer]

If the reports are all scheduled, the only problem with file-based events is if the timing of the event and the schedule get out of synch. For instance, if the process that creates the event file happens to complete before the scheduled time of a report, the report will not run; the event has to occur AFTER the report is scheduled to run. Of course, this can be remedied by renaming/deleting the file and renaming/creating it manually, giving the event server enough time to find the file gone before finding it present again.

If the users need to run the reports manually as well, they will either need to wait for the file event or you will need to write csp code to bypass the event requirement.

Another challenge is that event requirements cannot be set in ePortfolio; they have to be set in the CMC. And, if you have to change anything at the report level, it has to be done one report at a time. Crystal never made any provision for bulk changes to report-level (or user-level) attributes in Crystal Enterprise. I would hate to have to maintain over 100 reports; I think it would become a full-time job.

 

[spedi]

We have created a artificial time lines for each refresh/backup and those events never occur before.

As we use SQL Server databases, we have created a DTS package for each file event; Say the reports need to go out for server A is defined as 3:00 am then the reports are scheduled to run before 3:00 am and the file creation will never happen before as the file creation process is controlled by the DTS package. It will recreate the file only at 3:00 am or when ever the backup completes, which ever is later.

As far as the scheduling the report to use a event through eportfolio, I think we have overcome this, as we use the APOS tool that gives the access to events through eportfolio. Hence, our end users can use the events to run their our reports. I think you can integrate the events and reschedule feature for free using their tools. The tools are located at:

http://www.apos.on.ca/Freeware/

Enjoy

Srinath

http://www.nhsonline.org

srinath_p (at) yahoo.com

 

[MJRBIM]

The APOS InstanceManager tool lets you PAUSE pending instances in a BULK method...you might want to look into it as a solution.