Execution os script failed. reload old data? but with no indication of whats' wrong

[#1]

I am extracting data from 4 tables, 3 of the tables are liked correctly and with one partially linked the extract runs fine.

If I add the other link required which means changing the script to us the 'AS' function to just rename the required field I get the above error, but no indication in the "Script Execution Progess" box.

How do I switch on DEBUG mode, or how is the best way to identiry what's going wrong.

[#2]

[View:http://community.qlik.com/cfs-file.ashx/__key/CommunityServer.Discussions.Components.Files/11/7331.datamodel.doc:550:0]

If I remove the "as employee_number" the script runs without error

I am using version 9.00.7119.4, I can download an earlier version if you want me to try.

[Anonymous]

Clear enough, thanks.
You have empoyee_number field in two tables already - pydept and pyhistm_hpc. When you add a field with the same name to table wip_hpc, you create logical loop. You cannot use this field name here in your current logical design.
(No need to try another QV. I know that in QV 8.01 and earlier you'll be able to reload it, but QV will make a table or two "loosely coupled". QV 9, and probably 8.50 simply fail.)

[#4]

So I guess I need to remove the pydeptm table and any info I need from that extract with the pyhistm_hpc data as I only need description from pydeptm

[Anonymous]

I think removing pydempt is not a solution. You'll get:
employee_number in pyhistm_hpc and wip_hpc
year_period in pyhistm_hpc and and fin_cal
fin_date in fin_cal and wip_hpc
So, one of these must have different field name

[#6]

Michael not quite sure what you are saying now, problem is I have the employee name is in three tables which is needed along with other fields to make unique not sure what to do now.

Steve

[Anonymous]

Steve,
Here is what you have, if renaming to "employee_number".
1. Table pyhistm_hpc linked to table wip_hpc by employee_number.
2. Table pyhistm_hpc linked to table fin_cal by year_period.
3. Table wip_hpc linked to table fin_cal by fin_date.
You can see that it is a logical loop - circular reference. QV 9 reload fails for exactly this reason.
To be able to reload, you have to break the loop in any place. If you must have empoyee_number in both pyhistm_hpc and wip_hpc, you have to break this loop by renaming either year_period or fin_date in any table.
Another approach is to change your data model, maybe by joining two of the three tables in that loop. Maybe even all three. But this is defined mostly by your business needs, so I'm not in a good position to tell you what approach is the best in this case. Try and see.

[#8]

Michael

Got it sorted I think, I have attached my data model, but basically taken the data from pypersonm and pydeptm and loaded it with pyhistm and loaded fin_cal in with wip_hpc.

Just need to sort out the synthetic links now.

thanks Steve