alinaliBQ commented on issue #49404: URL: https://github.com/apache/arrow/issues/49404#issuecomment-4094289434
@amoeba @kou Regards to GH-49537, I checked and it appears that `cpack` is not designed to run using build artifact materials produced on another machine. `cpack` relies on hard-coded paths such as paths to WiX files within the repository. Instead of uploading artifact materials, we just upload the ODBC DLL arrow_flight_sql_odbc.dll` . To expand on kou's point, may I suggest this approach for `07-flightsqlodbc-upload.sh`: 1. [get ODBC DLL] Trigger CI A - This CI will build `arrow_flight_sql_odbc.dll` and upload the DLL to GitHub release as `arrow_flight_sql_odbc_unsigned.dll` to differentiate from signed DLL. 2. [sign ODBC DLL] Sign the ODBC DLL on local machine, then upload signed DLL to GitHub release as `arrow_flight_sql_odbc.dll`. 3. [get ODBC MSI] Trigger CI B - This CI will generate cpack files (and unsigned ODBC), then download signed `arrow_flight_sql_odbc.dll` from GitHub release, and then replace unsigned DLL with the signed DLL. Then run `cpack` and build the unsigned MSI and upload the MSI to GitHub release. This MSI will pick up the signed DLL. 4. [sign ODBC MSI] Sign the MSI on local machine, then upload signed MSI to GitHub release. If it sound good, I can work on the CI A and B needed in step 1 and 3. Please let me know what you guys think. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
