ryotaro2017 opened a new issue, #41905: URL: https://github.com/apache/superset/issues/41905
Graph app task: fe971df6-b138-491d-8803-b3d38b6a0aac Graph app detail page: http://localhost:8080/tasks/fe971df6-b138-491d-8803-b3d38b6a0aac ## Summary Superset defaults to a local directory for uploads (UPLOAD_FOLDER = BASE_DIR + "/static/uploads/"). In a horizontally scaled deployment, a file uploaded to one replica is not accessible on others, causing failures in file-serving endpoints and Hive CSV upload processing. ## Evidence - superset/config.py:1304 sets UPLOAD_FOLDER = BASE_DIR + "/static/uploads/" - superset/initialization/__init__.py:1039-1041 creates the directory on startup - superset/db_engine_specs/hive.py:254 uses it as the directory for temporary Parquet files during CSV-to-Hive uploads ## Proposed fix Abstract file storage behind an object-storage backend (S3/GCS/Azure Blob) or a documented shared network volume; introduce UPLOAD_STORAGE_BACKEND config. ## Proposed instrumentation Add a metric/health check that writes a test file from one replica and asserts it is readable from others; log upload folder path and replica identity. ## Scale-out impact This finding blocks horizontal scale-out for any feature that relies on the upload directory. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
