Hi, We would like to do a migration of data from some huge text files to the Mendix database. We are thinking of different options: Do the calls externally, transform every record from the file to an API request, so that Mendix has the control on how to store things Load the files into Mendix and process them record by record We believe both options will take time to load, and our architects are asking if the Mendix database (wich in this case is an on-premise SQL server instance) could be manipulated directly to insert all these information into the right tables. Initially, we discouraged this option as it could lead to integrity problems and because we could mess it up with the relationships, and especially with the generation of the ids. Also, we don’t know if the ids are generated as auto-numbers from the DB or if they’re generated by Mendix. However, for the comfort of everyone, we would like to check with the community in order to know: Is this technically feasible? Would you suggest this? Any other suggestions? Thanks, Francisco
Francisco Martínez Gallego
Even though the answer is six years old, I believe this thread it still relevant, and will answer your question. Summary: you shouldn't do this, because generating id's for new objects is very hard and will probably break your database.