After attempting an atomic host upgrade ( see details below ), docker0 disappeared. Chaos ensued.
Also, when rolling back, kubernetes seems not to trust that there is really space left on the device, and won't deploy containers to it.
So, effectively, atomic host upgrade took this node completely out of production. Since there happens to be persistence on this node ( volume mounts ), and the node (and only this node) is tagged to run the container that uses those volume mounts, I'm unable to start the service anywhere in the cluster until I go set up a new kubernetes node with the appropriate volume mounts.
I realize that it's on me to have configured it that way. ( We are early in R&D stages with Atomic Host. ) I mention those details because someone might find it interesting. I don't think my use-cases are terribly out-of-whack, just that the containers are not entirely configured like "cattle", in these early stages.
$ sudo atomic host status State: idle Deployments: ● centos-atomic-host:centos-atomic-host/7/x86_64/standard Version: 7.20160818 (2016-08-18 14:28:07) Commit: 709bfd3c4d9274c32bc3b691d642a3f95431136ec1a6b32e4cd5a13764647ba1 OSName: centos-atomic-host GPGSignature: 1 signature Signature made Thu 18 Aug 2016 09:39:07 AM CDT using RSA key ID F17E745691BA8335 Good signature from "CentOS Atomic SIG <firstname.lastname@example.org>" centos-atomic-host:centos-atomic-host/7/x86_64/standard Version: 7.20160224 (2016-02-24 08:01:24) Commit: 378538275fd9029a92435dc8c1efdb7a3486b04de47a2e6ac96752015b8352e5 OSName: centos-atomic-host GPGSignature: 1 signature Signature made Wed 24 Feb 2016 02:19:00 AM CST using RSA key ID F17E745691BA8335 Good signature from "CentOS Atomic SIG <email@example.com>"
Here are the entries from /var/log/messages:
Aug 30 07:01:51 aus-ared-dev-00 systemd: Starting Docker Storage Setup... Aug 30 07:01:51 aus-ared-dev-00 docker-storage-setup: INFO: Found an already configured thin pool /dev/mapper/cah-docker--pool in /etc/sysconfig/docker-storage Aug 30 07:01:51 aus-ared-dev-00 docker-storage-setup: INFO: Device node /dev/mapper/cah-docker--pool exists. Aug 30 07:01:51 aus-ared-dev-00 docker-storage-setup: Logical volume "docker-pool" changed. Aug 30 07:01:51 aus-ared-dev-00 systemd: Started Docker Storage Setup. Aug 30 07:01:51 aus-ared-dev-00 systemd: Starting Docker Application Container Engine... Aug 30 07:01:52 aus-ared-dev-00 ip: Cannot find device "docker0" Aug 30 07:01:52 aus-ared-dev-00 docker: flag provided but not defined: -d Aug 30 07:01:52 aus-ared-dev-00 docker: See '/usr/bin/docker-current --help'. Aug 30 07:01:52 aus-ared-dev-00 systemd: docker.service: main process exited, code=exited, status=125/n/a Aug 30 07:01:52 aus-ared-dev-00 systemd: Failed to start Docker Application Container Engine.
--- EDIT: Meant to add what happens when attempting rollback:
$ sudo atomic host rollback Moving '378538275fd9029a92435dc8c1efdb7a3486b04de47a2e6ac96752015b8352e5.0' to be first deployment Transaction complete; bootconfig swap: yes deployment count change: 0 Changed: NetworkManager 1:1.0.6-30.el7_2 -> 1:1.0.6-27.el7 NetworkManager-libnm 1:1.0.6-30.el7_2 -> 1:1.0.6-27.el7 atomic 1:1.10.5-7.el7 -> 1.6-6.gitca1e384.el7 bash 4.2.46-20.el7_2 -> 4.2.46-19.el7 bind-libs 32:9.9.4-29.el7_2.3 -> 32:9.9.4-29.el7_2.2 ...