Error delete _state: [403] Resource not accessible by integration

[autoantwort]

Description:
If a job has multiple actions/stale steps the creation of the cache fails.

Action version:
9

Platform:

• Ubuntu
• macOS
• Windows

Runner type:

• Hosted
• Self-hosted

Repro steps:
https://github.com/microsoft/vcpkg/blob/master/.github/workflows/stale.yml

Expected behavior:
No errors.

Actual behavior:
Errors. See https://github.com/microsoft/vcpkg/actions/runs/7750221792

[v-HarithaVattikuti]

Hello @autoantwort
Thank you for creating this issue. We will investigate it and get back to you as soon as we have some feedback.

[ericLemanissier]

Hello, the same error happens in here : https://github.com/conan-io/conan-center-index/actions/runs/7810903828/job/21305055330#step:2:1032

[sophies927]

Hello, I am encountering the same error here: https://github.com/microsoft/onnxruntime/actions/runs/7917906948/job/21615103971#step:2:3615.

[Pil0tXia]

Hi everyone, I'm encountering the same issue at https://github.com/apache/eventmesh/actions/runs/8675497930/job/23788986036 and I hope this log would be helpful to you.

Maybe I would revert back to v8 and see if this problem still exists.

[shmishra99]

Granting actions:write to the stale workflow will allow it to manage issues cache as intended.

[jeremystretch]

This appears to be the same issue described in #1131.

[autoantwort]

Adding actions: write also fixes parts of the problem. The same cache entry is then still used across different cache steps