# Out-of-memory file

Use with [When the data job runs out of memory](https://www.factualminds.com/blog/large-file-processing-glue-emr-athena-duckdb/). Start from the error. Checked 4 October 2026.

A 50 GB or 100 GB CSV loaded as one object is the case. Adding workers does not split a file the reader still opens whole.

| Question | Answer |
| --- | --- |
| File size and format | |
| Shape: interactive SQL / one-shot conversion / repeating pipeline | |
| Does the reader open the whole object? | |
| Choice: Athena / DuckDB on one node / Glue or EMR after the file is split | |

## Stop

- Another EMR deployment-mode matrix. That compare already exists.
- A compaction or catalog argument. Those posts already exist.