[Bug]: Traces not cleaned up after successful tests when using retain-on-failure, causing disk space issues v2.0

[TiagoMAGarcia]

Version

1.56.1

Steps to reproduce

To reproduce you just need to activate the "retain-on-failure" trace option and run tests. We have 4 workers and 972 tests. Here's some screenshots of the drive when the pipeline is running our tests:

Before the pipeline run:

10 minutes in:

20 minutes in:

Expected behavior

Traces for successful tests should be cleaned up immediately after each test completes successfully, rather than waiting until the end of the entire run. This would prevent disk space accumulation during long test runs.

Actual behavior

When using trace: 'retain-on-failure', traces for successful tests appear to only be cleaned up after the entire pipeline run completes. In our pipeline, this causes the dedicated drive to fill up during execution, leading to pipeline failures before all tests can complete.

Additional context

Related to: #38348

Environment

System:
    OS: Windows 11 10.0.22631
    CPU: (16) x64 AMD Ryzen 7 9800X3D 8-Core Processor           
    Memory: 1.59 GB / 31.10 GB
  Binaries:
    Node: 22.19.0 - C:\nvm4w\nodejs\node.EXE
    npm: 11.5.2 - C:\nvm4w\nodejs\npm.CMD
  IDEs:
    VSCode: 1.99.0 - D:\Microsoft VS Code\bin\code.CMD
  npmPackages:
    @playwright/test: 1.56.1 => 1.56.1 
    playwright: 1.56.1 => 1.56.1

[dgozman]

@TiagoMAGarcia This does not reproduce for us on a simple test, so you would need to share a full repro. Perhaps it is specific to your setup, so sharing your config and tests is essential.

Additionally, if you could investigate at which exact location the files that take up so much space are being stored, that would help a lot in pinpointing the problem. Thank you!