Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Use NamedTempFile rather than String in DiskManager #1680

Merged
merged 2 commits into from
Jan 26, 2022

Conversation

alamb
Copy link
Contributor

@alamb alamb commented Jan 25, 2022

Which issue does this PR close?

Closes #1679

Rationale for this change

  1. Using String for temporary files leaves the files around longer than necessary, and would cause trouble with a long lived DiskManager across plans.
  2. Using the existing tempfile crate in rust is likely to work better across operating systems than DataFusion specific tempfile creation logic

What changes are included in this PR?

  1. DiskManager passes out NamedTempFiles rather than String (when these are dropped they also clean up the temp file immediately)
  2. Update users to use NamedTempFiles

Are there any user-facing changes?

No

@alamb alamb added the api change Changes the API exposed to users of the crate label Jan 25, 2022
@alamb
Copy link
Contributor Author

alamb commented Jan 25, 2022

cc @yjshen

@github-actions github-actions bot added the datafusion Changes in the datafusion crate label Jan 25, 2022
path.push(file_name);
path.to_str().unwrap().to_string()
}
fn create_tmp_file(local_dirs: &[TempDir]) -> Result<NamedTempFile> {
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

tempfiles are now generated using tempfile rather than string manipulation

@@ -301,17 +306,16 @@ async fn spill_partial_sorted_stream(
}

fn read_spill_as_stream(
path: String,
path: NamedTempFile,
Copy link
Contributor Author

@alamb alamb Jan 25, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Since ownership of NamedTempFile is passed into the actual task doing the reading, when it is done, the temp file is cleaned up 🧹

Copy link
Contributor

@liukun4515 liukun4515 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nice work and great explanation
LGTM

@alamb alamb merged commit bffa5e4 into apache:master Jan 26, 2022
@alamb alamb removed the api change Changes the API exposed to users of the crate label Feb 10, 2022
@alamb alamb deleted the alamb/clean_up_tempfiles branch August 8, 2023 20:12
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
datafusion Changes in the datafusion crate
Projects
None yet
Development

Successfully merging this pull request may close these issues.

DiskManager keeps temporary files around until the manager itself is dropped
4 participants