Comment 2 for bug 1657173

Looking at the journal of any of the failing jobs, it seems that sshd is indeed started, but systemd is unable to understand it.

For example I see this:

Jan 19 16:31:53.852710 tripleo-centos-7-tripleo-test-cloud-rh1-6780446 systemd[1]: Started OpenSSH Server Key Generation.

then

Jan 19 16:31:53.883921 tripleo-centos-7-tripleo-test-cloud-rh1-6780446 systemd[1]: PID file /var/run/sshd.pid not readable (yet?) after start.

and after some time

Jan 19 16:33:25.117502 tripleo-centos-7-tripleo-test-cloud-rh1-6780446 systemd[1]: sshd.service start operation timed out. Terminating.

So it seems that the service file is unable or forgetting to write pid file, systemd intepret this as a failure, tries to restart

Jan 19 16:36:19.867524 tripleo-centos-7-tripleo-test-cloud-rh1-6780446 systemd[1]: sshd.service holdoff time over, scheduling restart.

then this happens

Jan 19 16:36:19.881772 tripleo-centos-7-tripleo-test-cloud-rh1-6780446 sshd[7345]: error: Bind to port 22 on 0.0.0.0 failed: Address already in use.

because sshd is really up and running.