This content has been marked as final. Show 4 replies
On the Import Analysis report, it is the Import: Process Map step that is taking 50% of the time.
FDM provides a breakdown of run-times by dimension - have you confirmed this is the dimension that is causing the issue?1 person found this helpful
when importing the file, what are the likely number of occurances of varvalues(19) = "SMM". If there are only a few, It might be better to test this variable first rather than the category key. It might also be worth exporing the possibility of combining the two tests with an 'and' to get rid of the nested if statements, to see if it provides any improvement.
If the same file is run for diffferent catkeys, are the results comparible or is the import for CatKey = 14 significantly slower?
Yes this dimension is one of those who provoke the issue. I'm testing your advices, on other hand the Import Performance Analysis shows this:
Process Description Duration I/O Source Event Information Date Start Time End Time
Import: Post Work Data 0 tDataMapSeg3 Post Map Work Data to Main Map Segment: 11/06/2012 9:31:08 AM 9:31:08 AM
Import: Post Work Data 7 tDataSeg3 Post Work Data to Main Data Segment: 11/06/2012 9:32:33 AM 9:32:40 AM
Import: Drop Work Table 5 tWbk0883870094684335 Drop Work Table 11/06/2012 9:34:01 AM 9:34:06 AM
What happening between "Import: Post Work Data" and "Import: Drop Work Table" there are temporal holes!
The first "Import: Post Work Data" ends at 9:31:08 AM and second "Post Work Data" starts at 9:32:33 AM. --> temporal hole 1:25
The second "Import: Post Work Data" ends at 9:32:40 AM and "Import: Drop Work Table" starts at 9:34:01 AM. --> temporal hole 1:21
Many Thanks !!
If you look at chapter 11 of the FDM API guide, you can see the actual steps the process takes, so you will be able to identify what takes place between the actions you have detailed.