build/release.sh script complains docker container name already assigned #1853
Comments
|
I've seen this happen but I thought I squashed it. One case that confused this stuff is if you are running non-interactive. An idea here is to:
I'm actually not hitting this at all right now so any detail of how and where might help us find a better way to work around this. |
|
I haven't seen this in quite a while. I'm going to close this. If you (or anyone else) starts seeing this again let me know and we can look at taking action. |
|
I hit this pretty consistently with "make release" on a fresh clone on my desktop. I'll send a PR later today (baby duty calls). |
|
Closed with #2780 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
When running build/release.sh, the kube::build::run_build_command removes any left over container from previous run, runs the container and then removes the container.
On faster systems, there could be a timing issue after the container has done running and the "docker rm" is done. This would cause the container not to be removed properly and create trouble building the container for the next arch.
I was able to fix this by introducing a "sleep 1" statement before the second "docker rm". There is a comment that says the ideal way is to use the "--rm" flag of docker run, but it is not done due to some known issue. Until that is resolved, if somebody runs into this issue, please try this workaround.
The text was updated successfully, but these errors were encountered: