Hide Forgot
Description of problem: After upgrading a RHEL host running docker 1.9 with images available on the host I can no longer run the original images or build new images. Version-Release number of selected component (if applicable): prior to upgrade # rpm -qa | grep docker docker-selinux-1.9.1-25.el7.x86_64 docker-forward-journald-1.9.1-25.el7.x86_64 docker-1.9.1-25.el7.x86_64 How reproducible: 100% Steps to Reproduce: 1. I installed docker 1.10 directly from brew, all binary rpms. yum install http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-common-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-forward-journald-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-logrotate-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-lvm-plugin-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-novolume-plugin-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-rhel-push-plugin-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-selinux-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-unit-test-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/docker-v1.10-migrator-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/oci-register-machine-1.10.3-27.el7.x86_64.rpm http://download.eng.bos.redhat.com/brewroot/packages/docker/1.10.3/27.el7/x86_64/oci-systemd-hook-1.10.3-27.el7.x86_64.rpm 2. Example running an image available prior to upgrade, all attempts to run or uild images seem to hit the same issue. docker run --rm -it registry.access.redhat.com/rhel7.2docker: Error response from daemon: Cannot start container 93b04b32f732cf32f447d3ebc847c53f1639b1d7016d4e51c908dd2798a66b45: [9] System error: exit status 1. Actual results: containers did not run our build Expected results: able to run existing pulled docker images on the host after upgrade to 1.10 Additional info:
Created attachment 1163773 [details] journalctl -b -u docker journalctl -b -u docker, I trimmed the log to show the start of docker after the upgrade
This looks like the unit file is not setup correctly by docker to use the cgroup configs?
Lokesh is docker using systemd for its cgroups?
I need to add --exec-opt native.cgroupdriver=systemd in the docker unitfile. This was already present in the docker-latest unitfile.
Verified I can run and build against existing and newly pulled images # rpm -q docker docker-1.10.3-28.el7.x86_64 # docker -v Docker version 1.10.3, build 4779225/1.10.3 # docker run --rm -it registry.access.redhat.com/rhel7.2 [root@98b256c29a0b /]# ls bin dev home lib64 mnt proc run srv tmp var boot etc lib media opt root sbin sys usr [root@98b256c29a0b /]# exit
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2016:1274