$ rpm -q containers-common
containers-common-1.0.0-1.fc32.x86_64
$ podman image build ...
Building redmine image
STEP 1: FROM ubuntu:xenial-20180705 AS ubuntu-xenial-20180705
Error: error creating build container: (image name "ubuntu:xenial-20180705" is a short name and no search registries are defined in /etc/containers/registries.conf): while pulling "ubuntu:xenial-20180705" as "localhost/ubuntu:xenial-20180705": Error initializing source docker://localhost/ubuntu:xenial-20180705: error pinging docker registry localhost: Get "https://localhost/v2/": dial tcp [::1]:443: connect: connection refused
I see that the update already has 3 karma, rather quickly in 2 days: is no one else seeing this issue apart from me? Has the Dockerfile format that podman uses changed (in which case this is a major update and should not perhaps be pushed to stable releases now, right?)
$ git clone https://github.com/SilverLabUCL/docker-redmine-osb.git
$ cd docker-redmine-osb
$ podman build --tag "redmine" -f Dockerfile
STEP 1: FROM ubuntu:xenial-20180705 AS add-apt-repositories
Error: error creating build container: (image name "ubuntu:xenial-20180705" is a short name and no search registries are defined in /etc/containers/registries.conf): while pulling "ubuntu:xenial-20180705" as "localhost/ubuntu:xenial-20180705": Error initializing source docker://localhost/ubuntu:xenial-20180705: error pinging docker registry localhost: Get "https://localhost/v2/": dial tcp [::1]:443: connect: connection refused
OK, I'm confused. Is this an issue with Podman or Skopeo? All the problems in the comments are for Podman build issues.... Could I ask that you create an upstream issue against https://github.com/containers/libpod/issues please?
The issue is clearly with this update, since podman has not been updated on our systems. If this update requires podman version 2 from the other update, then the two builds need to be pushed together as one update to prevent breakages.
Our of curiosity: what version of podman are you on that you did (or are not) seeing this bug?
@ankursinha, thanks for lifting my confusion, I'm on the same page now. You had Podman installed on the system prior, you installed this skopeo rpm, and then Podman went belly up. Thanks for the explanation.
This update's test gating status has been changed to 'passed'.
So, can someone confirm if this is a general bug? @tomsweeneyredhat, @dwalsh, @jwhonce: could you provide more info please? What tests did you do where you did not encounter this issue? What version of podman did you use?
Wouldn't all of us with a prior podman installation see this bug, and so won't all users now run into it?
@ankursinha this build made it out by accident, but it was caught quickly and replaced by the -2 build. The bug only affected shortnames ("podman pull foo", vs "podman pull registry/name/foo").
This update has been submitted for testing by lsm5.
This update's test gating status has been changed to 'waiting'.
This update's test gating status has been changed to 'ignored'.
This update has been pushed to testing.
This update's test gating status has been changed to 'passed'.
LGTM
lgtm
This update can be pushed to stable now if the maintainer wishes
This update has been submitted for stable by rhcontainerbot.
LGTM and passes rudimentary testing.
This update's test gating status has been changed to 'greenwave_failed'.
Hrm, seeing a regression here, but not sure what's causing it:
With the previous version, I was able to build images without any issue:
After the update, this doesn't work any more:
I see that the update already has 3 karma, rather quickly in 2 days: is no one else seeing this issue apart from me? Has the Dockerfile format that podman uses changed (in which case this is a major update and should not perhaps be pushed to stable releases now, right?)
The Dockerfile is in the repository here: https://github.com/SilverLabUCL/docker-redmine-osb/blob/master/Dockerfile
So, to reproduce:
Works fine after downgrading.
Is podman v2 required for this? I've got 1.9.3 here: podman-1.9.3-1.fc32.x86_64
The podman v2 update for F32 hasn't even hit testing yet, and has 0 karma so it won't get to stable any time soon either: FEDORA-2020-6660e9ee6e
Confirmed. This containers-common has a bad registries.conf.
OK, I'm confused. Is this an issue with Podman or Skopeo? All the problems in the comments are for Podman build issues.... Could I ask that you create an upstream issue against https://github.com/containers/libpod/issues please?
The issue is clearly with this update, since podman has not been updated on our systems. If this update requires podman version 2 from the other update, then the two builds need to be pushed together as one update to prevent breakages.
Our of curiosity: what version of podman are you on that you did (or are not) seeing this bug?
@tomsweeneyredhat /etc/containers/registries.conf comes from containers-common rpm which == skopeo
@ankursinha, thanks for lifting my confusion, I'm on the same page now. You had Podman installed on the system prior, you installed this skopeo rpm, and then Podman went belly up. Thanks for the explanation.
This update's test gating status has been changed to 'passed'.
This update has been pushed to stable.
So, can someone confirm if this is a general bug? @tomsweeneyredhat, @dwalsh, @jwhonce: could you provide more info please? What tests did you do where you did not encounter this issue? What version of podman did you use?
Wouldn't all of us with a prior podman installation see this bug, and so won't all users now run into it?
@ankursinha this build made it out by accident, but it was caught quickly and replaced by the -2 build. The bug only affected shortnames ("podman pull foo", vs "podman pull registry/name/foo").
Ah, thanks. I'll go test that one out now. (tagging doesn't seem to send notifications on bodhi or I'd have seen it before :( )