Overview

When using runAsNonRootUser and runAsUser you may experience the following error message in your CircleCI job.

CircleCI failed to run this build, check your config. Try re-running the build and 
if this issue persists, open a Support ticket. Detail: failed to copy circleci-agent 
into container &#34;primary&#34;, 
Error: error executing command mkdir -p $HOME/.local/bin/circleci : 
error executing command /bin/bash -c set -o pipefail ; mkdir -p 
$HOME/.local/bin/circleci 2&gt;&amp;1 | tee /proc/1/fd/1 : 
command terminated with exit code 1

This is typically a case of the user in runAsUser not matching the User ID within the image.

Solution

Step 1

Verify the User ID configured in the Docker image.

$ docker pull ${image_name}
$ docker run --rm ${image_name} echo $UID

Step 2

Confirm the runAsUser matches the UID returned by the first command.

agent:
  resourceClasses:
    <namespace>/<runner_name>:
      spec:
        containers:
        securityContext:
          runAsNonRoot: true
          runAsUser: <the_UID_found_before>

Example

As an example, we can test this with a cimg/base:current image.

$ docker pull cimg/base:current
$ docker run --rm cimg/base:current echo $UID
  1000

We can see that after running the docker run command, it returns 1000 to us.

If your runAsUser does not match the same as $UID in the Docker image, you need to update either one of them to allow your container to run in your Container Runner.

Additional Notes

Container Runner job failing to copy circleci-agent into container