Skip to main content

When reading the documentation about preparing a migration job that uses procedure Output File, the text shows:

Write data to a server file

When File Location = OnServer File Path must be a valid directory object. The column has a List-Of-Value with valid directory objects.

So logged in as IFSAPP in the database an Oracle directory MIGOUTFILES was created with C:\MigOutFiles as path on the server (I have created it on the database server).

Let's take the company as view to export. I select MIGOUTFILES as output path. File name is Companies.txt and File location is set to OnServer.

To make this as simple as possible, in the File Mapping only one field/row is created with field COMPANY.

All rules are inactive.

When starting this migration job online the system reports: Failed to open file Companies.txt on MIGOUTFILES.

The windows directory exist, so why do I get this message?

Hi @eqbstal,

Could you try to create the directory logged in as SYS and grant READ, WRITE  permissions to the directory as shown below?

CREATE OR REPLACE DIRECTORY MIGOUTFILES AS ‘C:\MigOutFiles’;

GRANT READ, WRITE ON DIRECTORY MIGOUTFILES TO <AppOwner>;

Hope this helps!

Thanks @Charith Epitawatta for the idea. I'm currently trying to locate the person who knows the password of SYS. It took me already 10 cups of coffee 😄

@Charith Epitawatta Yes, this was the solution. A real pity that this is not in the documentation of migration jobs.

Thanks for sharing your knowledge.

++ Additional Infor 


When creating the directory path, ensure you double-check the physical path. It can either be located on the database server itself or on a shared path—both options are valid. If using a shared path, ensure it is specified correctly and accessible from the database server.

 

Thanks,

Ashen

Reply


Terms & ConditionsCookie settings