Github/actions: GetError: Can't use 'tar -xzf' extract archive file return code: 2
bdoublet91 opened this issue · comments
Hi,
I use github action with self hosted runners and I don't get problem until now.
name: CI mirroring
# Controls when the action will run.
on:
push:
pull_request:
schedule:
# * is a special character in YAML so you have to quote this string
- cron: '30 2 * * *'
# A workflow run is made up of one or more jobs that can run sequentially or in parallel
jobs:
# This workflow contains a single job called "build"
build:
# The type of runner that the job will run on
runs-on: self-hosted
timeout-minutes: 15
# Steps represent a sequence of tasks that will be executed as part of the job
steps:
# Checks-out your repository under $GITHUB_WORKSPACE, so your job can access it
- uses: actions/checkout@v2
I want to checkout my private github repo to execute action but get error that I haven't before:
Current runner version: '2.278.0'
Runner name: 'runner_1'
Runner group name: 'Default'
Machine name: '2526d86f2a75'
GITHUB_TOKEN Permissions
Prepare workflow directory
Prepare all required actions
Getting action download info
Download action repository 'actions/checkout@v2'
Error: Can't use 'tar -xzf' extract archive file: /tmp/runner/_actions/_temp_0207f89a-ccec-4204-a7b6-533817e1094e/cef27ea4-0563-40a2-ac70-c94e38e49e4d.tar.gz. return code: 2.
When I use github runner with ubuntu-latest, it's working so I think there is something wrong with my self hosted runner.
I use docker self hosted runner from this project https://github.com/myoung34/docker-github-actions-runner
All working well, runners start, connect to github and get idle state.
When I execute pipeline, runner take the job and execute but get this error. I don't know what change.
I have tar binary and the error means that tar was executed but failed so maybe the download of tar.gz file failed.
which tar
/usr/bin/tar
Can you help me to debug this ? debug logs, some checks, any idea.
Thanks you
Best regards
Return code 2 for tar means there was a file read error. You'll want to check permissions for the file and extract path
Thanks for your quick answer !
yeah, I read this on tar documentation.
here my workdir_runner:
ls -ahl /tmp/runner/_actions/
total 12K
drwxr-xr-x 3 nobody nogroup 4.0K May 26 21:27 .
drwxr-xr-x 6 nobody nogroup 4.0K May 26 21:27 ..
drwxr-xr-x 3 nobody nogroup 4.0K May 26 21:27 actions
seems like folder are created because I have Folder with my repo name but there is no tar.gz file.
Could be deleted after the job failed ? or doesnt exist ?
Do you know what is this tar.gz file ?
There is some command I can reproduce to test the download of tar.gz file manually ?
This is not something I'm familiar with, the tar usage I've had has had no issues locally or in testing
Ok I'll dig this problem and do a feedback it I find a solution.
Thanks
OK found it.
volumes:
- '/tmp/runner:/tmp/runner'
# needs to be the same path on host and inside the container,
I thought this setting was necessary if you use docker command in github-runner. So I set up the host path and container path differently and this error occured. So I rollback and it's ok now.
Just a question ? If I don't use docker command, why did this happen ? I mean, why I can't use different path for host and container for workdir runner ?
The issue can be closed. Thanks
Just a question ? If I don't use docker command, why did this happen ? I mean, why I can't use different path for host and container for workdir runner ?
I think this is internal to the way github runners work outside their infra more than it is docker, but I'm not completely sure.
Ok, anyway, thanks you for your job about this project.
This started happening for us this morning too. No changes in code between the working and breaking state
Started happening to us in the past couple of hours too.
Although we're using a different self hosted runner.
The same is happening to us, we aren't even using hosted runners
We're having the same issue, it fails sometimes on all our workflows, using the github hosted runners
Same error here. Randomly failing test runs with the error given in the issue. Restarting the run might help, but not always.
Same error here also 👋
Bump
Same problem. It happened since this morning. 🥲
actions/runner#1923 seems like a better place to track this issue, though.
raised issue : actions/runner#1923
Still happening here without any changes on our end
It happens randomly for me (for a few CI runs this morning)
looks like it's working now..
Thanks for all the quick info. Looks like something upstream was definitely broken. It's worth leaving feedback on actions/runner#1923