Comments

65 Comments

Great. I can verify my scratch-builds on f28 succeeded. Thanks for the override.

nss requites nss-util > 3.37.3. Given the nss is already pushed into the stable and this is not, all [scratch]-builds in f28 requiring nss currently fails. Started to happen like 20 minutes ago. Can anyone override the nss-util rpm for a couple of day before this update gets into the stable?

Can you override the build? So it gets faster into the buildroot. I am updating cloud.google.com/go to 0.21.0 and this is one of its dependencies on f28.

Can you override the build? So it gets faster into the buildroot. I am updating cloud.google.com/go to 0.21.0 and this is one of its dependencies on f28.

Even just with the docker-1.13.1-52.git89b0e65.fc29.x86_64 there are errors:

Apr 17 12:33:39 f28-k8s-test.localdomain dockerd-current[768]: container_linux.go:247: starting container process caused "process_linux.go:258: applying cgroup configuration for process caused \"No such device or address\""

I backported https://github.com/kubernetes/kubernetes/pull/61926 on top of 1.9.6 and even with the latest docker installed the kubelet still complains. Not sure if it the issue is still in docker or something is missing in Kubernetes.

I would love to but we are still stuck on https://bugzilla.redhat.com/show_bug.cgi?id=1409553. Otherwise we violate the Fedora packaging guidelines cause some Golang packages are already in RHEL and we can not update them in epel7.

We can't push this one as it conflicts with the same package in RHEL7

1468752 is fixes via kubernetes/contrib/init/systemd/environ/kubelet file. It works for a non-ansible deployment.

The unit-test-devel rpm is not for installing atm. It is for a CI that is to be implemented. To test the code we need environment that is hard or impossible to create inside the Koji builder.