Description of problem: Installing IPI on OSP with proxy, nodeip-configuration.service is not starting because podman is trying to download a container from quay.io using a direct connection instead of using the proxy: [root@mrnd-13-46-px-nvlt8-master-0 etc]# journalctl -u nodeip-configuration.service -- Logs begin at Fri 2020-06-26 09:00:21 UTC, end at Fri 2020-06-26 09:17:30 UTC. -- Jun 26 09:07:27 mrnd-13-46-px-nvlt8-master-0 systemd[1]: Starting Writes IP address configuration so that kubelet and crio services select a valid node IP... Jun 26 09:07:28 mrnd-13-46-px-nvlt8-master-0 podman[1531]: Trying to pull quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:e9163b67ae38a9d9be54acd42bc1e3496d46> Jun 26 09:08:21 mrnd-13-46-px-nvlt8-master-0 systemd[1]: nodeip-configuration.service: Main process exited, code=killed, status=15/TERM Jun 26 09:08:21 mrnd-13-46-px-nvlt8-master-0 systemd[1]: nodeip-configuration.service: Failed with result 'signal'. Jun 26 09:08:21 mrnd-13-46-px-nvlt8-master-0 systemd[1]: Stopped Writes IP address configuration so that kubelet and crio services select a valid node IP. Jun 26 09:08:21 mrnd-13-46-px-nvlt8-master-0 systemd[1]: nodeip-configuration.service: Consumed 266ms CPU time -- Reboot -- Jun 26 09:08:42 mrnd-13-46-px-nvlt8-master-0 systemd[1]: Starting Writes IP address configuration so that kubelet and crio services select a valid node IP... Jun 26 09:08:43 mrnd-13-46-px-nvlt8-master-0 podman[1253]: Trying to pull quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:e9163b67ae38a9d9be54acd42bc1e3496d46> Jun 26 09:09:43 mrnd-13-46-px-nvlt8-master-0 podman[1253]: Get https://quay.io/v2/: dial tcp 172.18.124.15:443: i/o timeout Jun 26 09:09:43 mrnd-13-46-px-nvlt8-master-0 podman[1253]: Error: unable to pull quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:e9163b67ae38a9d9be54acd42bc1e> Jun 26 09:09:43 mrnd-13-46-px-nvlt8-master-0 systemd[1]: nodeip-configuration.service: Main process exited, code=exited, status=125/n/a Jun 26 09:09:43 mrnd-13-46-px-nvlt8-master-0 systemd[1]: nodeip-configuration.service: Failed with result 'exit-code'. Jun 26 09:09:43 mrnd-13-46-px-nvlt8-master-0 systemd[1]: Failed to start Writes IP address configuration so that kubelet and crio services select a valid node IP. Jun 26 09:09:43 mrnd-13-46-px-nvlt8-master-0 systemd[1]: nodeip-configuration.service: Consumed 252ms CPU time Note: quay.io resolves to 172.18.124.15 because it is the way we can block direct connections to internet. Version-Release number of selected component (if applicable): 4.6.0-0.nightly-2020-06-26-035408 How reproducible: Steps to Reproduce: 1.Install using proxy on install-config.yaml proxy: httpProxy: http://10.0.79.123:3128 httpsProxy: http://10.0.79.123:3128 noProxy: 'rhos-d.infra.prod.upshift.rdu2.redhat.com,oauth-openshift.apps.mrnd-13-46-px.qe.devcluster.openshift.com' 2.Set DNS to block resolution on quay.io to catch any direct connection done by the node 3.Check status of nodeip-configuration.service Actual results: Service fails to start, it launch a container trying to download it directly from quay.io without using the proxy configured on the system Expected results: Service used the proxy configured on the system for any http connection Additional info:
The nodeip-configuration.service unit file doesn't export the environment variables to use the proxy. I'll submit a tentative patch shortly
Verified on 4.6.0-0.nightly-2020-07-10-092830
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 (OpenShift Container Platform 4.6 GA Images), 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-2020:4196