-
Notifications
You must be signed in to change notification settings - Fork 2.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Podman run oci-archive gives "Error: invalid reference format" #6744
Comments
Thanks for opening the issue @gatoniel! Podman can only create containers with images that are in the local container storage. This means, that it only supports the Is your use case to set up a storage that multiple users can share? We support the concept of @rhatdan you wrote a more recent blog about additional stores explicitly but I fail to find it. Has it been released? |
Hi,
Somehow this worked until some days ago. But then I will use a different setup.
Yes, that is my use case. I read the article and will try to reproduce this on my system by adding
to /etc/containers/storage.conf. But how do I now add images to the first store? Or would it also be safe, to just add my own user store? I think the workflow would be:
I want to avoid running sudo together with skopeo or podman, to be sure that everything I run can also be run by other normal users. But the skopeo statement does not work and throws this error:
|
Can you pull the image with |
I could, but this would mean rewriting some other code, since the podman run command is executed within a Python tornado server (JupyterHub with podman spawner). Another upside of the first mentioned solution with an additional storage is the disk usage, isnt it? When every user pulls the image, they need to save it in their own user storage under .local/share/containers . When I use the additional storage that would not be the case. This would save lots of disk space. Best, |
Please scratch what I said before:
For sure, Podman will pull images and given you report it has worked before, we may regressed here. I will investigate. |
I can reproduce, thanks again! |
But as I understand the |
Opened #6751 to fix the issue. |
Support all image transports in podman run/create. It seems we regressed with v2 on that. Also add tests to make sure we're not regressing again. Fixes: containers#6744 Signed-off-by: Valentin Rothberg <[email protected]>
If our additional store is located |
Soon, a new blog from @rhatdan will be released on Red Hat SysAdmin which will cover this topic in greater detail. |
Support all image transports in podman run/create. It seems we regressed with v2 on that. Also add tests to make sure we're not regressing again. Fixes: containers#6744 Signed-off-by: Valentin Rothberg <[email protected]>
had this error because of a trailing slash only, removed that and this error was gone :) |
Is this a BUG REPORT or FEATURE REQUEST? (leave only one on its own line)
/kind bug
Description
Hi,
I am supplying containers to users on a workstation. My current workflow is like this:
vim Dockerfile
podman build . -t myimage:1.0 --format=docker
podman push myimage:1.0 oci-archive:/home/myuser/images/myimage:1.0
sudo cp -r /home/myuser/images/myimage /mnt/datahdd/readonly/images/myimage
podman run oci-archive:/mnt/datahdd/readonly/images/myimage:1.0
I am happy to hear about a better workflow, but this worked for me until today. Now the
podman run
gives an error:Error: invalid reference format
. When I runpodman run myimage:1.0
everything works fine. Butpodman run oci-archive:/home/myuser/images/myimage:1.0
gives me the same error.I also tried pushing via
podman push myimage:1.0 oci-archive:/home/myuser/images:myimage:1.0
and running viapodman run oci-archive:/home/myuser/images:myimage:1.0
, but the error stays. I also could not resolve this issue by removing the--format=docker
statement in the build command.As I said, this error occured just recently and this workflow seemed to work before. The only thing I recently changed is that I set up the gpg-agent. Might this disturb some processes? I also do not know, if I need to initialize the oci-archive path beforehand.
Steps to reproduce the issue:
podman pull registry.fedoraproject.org/f29/httpd
podman push registry.fedoraproject.org/f29/httpd oci-archive:/home/netter/administration/podman-images/httpd
podman run oci-archive:/home/netter/administration/podman-images/httpd
Describe the results you received:
I get:
Describe the results you expected:
The container should run, as if I was using the command
podman run registry.fedoraproject.org/f29/httpd
Additional information you deem important (e.g. issue happens only occasionally):
I think the only thing I was changing in my setup was adding gpg-agent environment variables. In my envs I found this environment variable:
OLDPWD=/home/netter/administration/podman-images
Output of
podman version
:Output of
podman info --debug
:Package info (e.g. output of
rpm -q podman
orapt list podman
):Additional environment details (AWS, VirtualBox, physical, etc.):
It is a physical workstation running Centos 8.
The text was updated successfully, but these errors were encountered: