The docker startup script seems always happy. It's [ok], # /etc/init.d/docker restart * Stopping docker ... [ ok ] * Starting docker ... [ ok ] but something is wrong # docker images Cannot connect to the Docker daemon. Is the docker daemon running on this host? grep shows no process, /var/log/docker.log reveals some problem time="2016-01-31T15:08:26.523265597+01:00" level=fatal msg="Error starting daemon: error initializing graphdriver: Base Device UUID and Filesystem verification failed.Current Base Device UUID:4181ff8d-fdb4-4d21-88f3-dda02e371a14 does not match with stored UUID:92b5b9a3-2d92-42a0-a967-8dc525c5fc7f. Possibly using a different thin pool than last invocation" Now the concrete problem is not topic of this ticket, just the startup script bahavior. I'd expect "starting docker" to fail and not to show [ ok ] subbornly. Reproducible: Always
Same result when PaX prevents it from working. initscript shows "ok", but log contains the following messages and all docker commands simply wait forever: time="2016-08-13T13:40:54.939104136Z" level=info msg="New containerd process, pid: 3030\n" containerd: error while loading shared libraries: cannot make segment writable for relocation: Permission denied
Can you reproduce this with the latest >=docker-1.13.1-r1?
