TM1 Forum

Hi All,

I was curious to know if there is any way to resume a TI process once an error is thrown.
I know ProcessBreak function will directly take the process to the Epilog Tab. But is there a way to resume a TI process (for example, in Metadata/Data tab once an error is thrown in the middle of the code there?)

Thanks !

If a process encounters a major error it will quit, this is the designed behaviour and there isn't anything you can do about that. There is no equivalent (as far as I am aware) of On Error Resume Next as in VB and VBA. You can increase the tolerance for minor errors by manipulating the variables OnMinorErrorDoItemSkip and MinorErrorLogMax on the prolog tab. The best way to ensure a process completes is to make sure that the datasource is clean and contains no errors, failing that then make sure all expected errors are handled within the code.

lotsaram wrote:If a process encounters a major error it will quit, this is the designed behaviour and there isn't anything you can do about that. There is no equivalent (as far as I am aware) of On Error Resume Next as in VB and VBA. You can increase the tolerance for minor errors by manipulating the variables OnMinorErrorDoItemSkip and MinorErrorLogMax on the prolog tab. The best way to ensure a process completes is to make sure that the datasource is clean and contains no errors, failing that then make sure all expected errors are handled within the code.

Hey Lotsa,
This is a slightly different question but we seem to notice that logging is constantly getting turned off for certain cubes. I can't exactly tell when but there are only 2 TI's that load the cube. End users can press an action button to load the data. Both TI's have the potential to have Data Errors (Generally do to missing items in one of the dimensions.) Data errors shouldn't prevent the TI from going to the Epilog and processing the portion of code to turn logging back on right?

PlanningDev wrote:[ This is a slightly different question but we seem to notice that logging is constantly getting turned off for certain cubes. I can't exactly tell when but there are only 2 TI's that load the cube. End users can press an action button to load the data. Both TI's have the potential to have Data Errors (Generally do to missing items in one of the dimensions.) Data errors shouldn't prevent the TI from going to the Epilog and processing the portion of code to turn logging back on right?

It is conceivable that on encountering a major error that the processes might quit and leave subsequent epilog code unexecuted. To minimize risk of unlogged transactions I always have a background process running every 10 or 15 minutes (during working hours only) that loops through the cubes and makes sure that logging is turned on.