Temp files remain in /tmp directory after successful file uploads that can leave device out of space

[ychoi0822]

Bug report

Describe the bug

After files are uploaded successfully to the final location, temp files remain on the /tmp directory. After a while, this results in no space in disk.

Steps to reproduce the behavior

1. Go to admin console.
2. Click on "media library"
3. Click on "upload files"
4. Check the files are successfully uploaded to the final location
5. Check the same files still exist in /tmp directory

Expected behavior

After file has been uploaded successfully to the final location, temp files are removed from /tmp directory.

Screenshots

System

• Node.js version: 14.15.4
• NPM version: 6.4.1
• Strapi version: 3.6.5
• Database: PostgresSQL
• Operating system: Linux

Additional context

This was raised previously here - #2752

[derrickmehaffy]

Was able to confirm on v3.6.5 on local and AWS-S3 provider

[fabiel-leon]

also using strapi-provider-upload-tp-minio on v3.6.5

[derrickmehaffy]

also using strapi-provider-upload-tp-minio on v3.6.5

This is an external package, we can't fix that.

[AAS97]

What is the status on that issue?
How are we supposed to manage on production servers?

[derrickmehaffy]

This issue has been mentioned on Strapi Community Forum. There might be relevant details there:

https://forum.strapi.io/t/temporary-large-disk-usage-crashing-strapi/12297/2

[wintercounter]

I have the same issue with sending files to a custom controller endpoint, unrelated to the upload plugin. I'm on 3.6.8, don't know if this exists in v4.

[adconk]

I'm using Strapi 4.5.1 and I get an error that is suspiciously similar to this. My Strapi container service will throw this error from time to time and restart: "Error: ENOSPC: no space left on device, write".

[derrickmehaffy]

I believe this should be fixed in v4. @adconk if you continue to get this issue can you open a new issue. Typically it shouldn't be caused by Strapi but maybe something else is writing into your tmp?