I have read everything I could get my hands on. We are attempting to update several contact records in the contact table (just one field), the issue is that there are many users without a middle_name value. The middle_name says in schema designer that it is "required for pdm_userload". We are having no issues including the middle_name as long as the user has a middle name, everything works fine. However, when there is no middle_name I get a "cannot have null value" error. Then when I try to remove the middle_name it from the .dat file and run it I get an invalid logical key error, skipping table. Is there a way to map a null value in the .dat file that will allow me to include the middle_name which is required? Do I need to do a deref for the middle_name value? If so, will that effect anything if it is null?
The term specifically in the Schema Designer is "Key for pdm_userload" which is the reason I had to include it. If it wasn't included = Logical Key Error; if included = "Cannot be Null" Error
I would recommend using pdm_load in place of pdm_user load. BUT, first the best thing to do is to do a pdm_extract of your entire ca_contact table, so that you can keep it as a backup in case you need to reference it later. Then you can just load in the .dat file using pdm_load. This should allow you to update the contacts that you want to. As long as your .dat file contains UUIDs then it will not create duplicates
Hope this helps,
As Jon Israel mention using pdm_load will bypass this. Backup strongly recommended
That say I always use Web service to upload data vs. pdm_load to avoid any risk of data corruption.
This worked perfectly in our test environment. We used UUID for the name was able to use other fields as well. Thank you very much!