The named docker container for the logging agent is still around, preventing the agent from starting after the VM is restarted.
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: stackdriver-logging.service: Service RestartSec=10s expired, scheduling restart.
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: stackdriver-logging.service: Scheduled restart job, restart counter is at 54.
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: Stopped Fluentd container for Stackdriver Logging.
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: stackdriver-logging.service: Consumed 58ms CPU time
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: Starting Fluentd container for Stackdriver Logging...
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: Started Fluentd container for Stackdriver Logging.
Oct 16 16:34:24 alpha-relay-kmrj docker[1742]: /usr/bin/docker: Error response from daemon: Conflict. The container name "/stackdriver-logging-agent" is already in use by container "7cf099e84bd1841100c8718df32f6bbb9c78897a0f42b7c2b1e0e92175711dd0". You have to remove (or rename) that container to be able to reuse that name.
Oct 16 16:34:24 alpha-relay-kmrj docker[1742]: See '/usr/bin/docker run --help'.
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: stackdriver-logging.service: Main process exited, code=exited, status=125/n/a
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: stackdriver-logging.service: Failed with result 'exit-code'.
Oct 16 16:34:24 alpha-relay-kmrj systemd[1]: stackdriver-logging.service: Consumed 56ms CPU time
The named docker container for the logging agent is still around, preventing the agent from starting after the VM is restarted.