Root cause analysis:

Symptoms: 
When the <Save and Quite> key pressed, the script hangs on systemctl start docker.
Actually the docker.service wait docker-storage storage creating LVM volume, which in his turn 
hang up on the semaphore.

Here the CLI investigation
=========================================================================
(gdb) bt
#0  0x00007f7ce75884e7 in semop () from /lib64/libc.so.6
#1  0x00007f7ce7cb40cc in dm_udev_wait () from /lib64/libdevmapper.so.1.02
#2  0x00007f7ce88537bd in fs_unlock ()
#3  0x00007f7ce8860a05 in _file_lock_resource ()
#4  0x00007f7ce87f3878 in _lock_vol ()
#5  0x00007f7ce87f41c3 in lock_vol ()
#6  0x00007f7ce8808b6c in wipe_lv ()
#7  0x00007f7ce880c54f in _lv_create_an_lv ()
#8  0x00007f7ce880d55f in lv_create_single ()
#9  0x00007f7ce879e13a in lvcreate ()
#10 0x00007f7ce87a63d8 in lvm_run_command ()
#11 0x00007f7ce87a6e40 in lvm2_main ()
#12 0x00007f7ce74b1b15 in __libc_start_main () from /lib64/libc.so.6
#13 0x00007f7ce878de41 in _start ()
(gdb)

# ipcs -a

------ Message Queues --------
key        msqid      owner      perms      used-bytes   messages    

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status      

------ Semaphore Arrays --------
key        semid      owner      perms      nsems     
0x0d4dbf3c 98304      root       600        1         

# ipcs -sti 98304

Semaphore Array semid=98304
uid=0	 gid=0	 cuid=0	 cgid=0
mode=0600, access_perms=0600
nsems = 1
otime = Wed Jul  6 12:29:04 2016  
ctime = Wed Jul  6 12:29:04 2016  
semnum     value      ncount     zcount     pid       
0                     1           0           0          25896     


# ps xawf -eo pid,ppid,user,cgroup,args

    1     0 root     -                           /usr/lib/systemd/systemd --switched-root --system --deserialize 21
  491     1 root     7:devices:/system.slice/sys /usr/lib/systemd/systemd-journald
  505     1 root     7:devices:/system.slice/lvm /usr/sbin/lvmetad -f
  512     1 root     7:devices:/system.slice/sys /usr/lib/systemd/systemd-udevd
  662     1 root     7:devices:/system.slice/aud /sbin/auditd -n
  686     1 root     7:devices:/system.slice/irq /usr/sbin/irqbalance --foreground
  687     1 dbus     7:devices:/system.slice/dbu /bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
  692     1 root     7:devices:/system.slice/sys /usr/lib/systemd/systemd-logind
  698     1 root     7:devices:/system.slice/cro /usr/sbin/crond -n
  715     1 root     7:devices:/system.slice/wpa /usr/sbin/wpa_supplicant -u -f /var/log/wpa_supplicant.log -c /etc/wpa_supplicant/wpa_supplicant.conf -u -f /var/log/wpa_supplica
  716     1 polkitd  7:devices:/system.slice/pol /usr/lib/polkit-1/polkitd --no-debug
10693     1 root     7:devices:/system.slice/rsy /usr/sbin/rsyslogd -n
11272     1 ntp      7:devices:/system.slice/ntp /usr/sbin/ntpd -u ntp:ntp -g
25796     1 root     7:devices:/system.slice,1:n /usr/sbin/sshd -D
26982 25796 root     1:name=systemd:/user.slice/  \_ sshd: root@pts/0
26986 26982 root     1:name=systemd:/user.slice/  |   \_ -bash
27011 26986 root     1:name=systemd:/user.slice/  |       \_ lvdisplay
27039 25796 root     1:name=systemd:/user.slice/  \_ sshd: root@pts/1
27043 27039 root     1:name=systemd:/user.slice/  |   \_ -bash
27279 27043 root     1:name=systemd:/user.slice/  |       \_ gdb attach 25896
29387 25796 root     1:name=systemd:/user.slice/  \_ sshd: root@pts/2
29391 29387 root     1:name=systemd:/user.slice/      \_ -bash
29543 29391 root     1:name=systemd:/user.slice/          \_ ps xawf -eo pid,ppid,user,cgroup,args
25861     1 root     7:devices:/system.slice,1:n /bin/bash /usr/bin/docker-storage-setup
25896 25861 root     7:devices:/system.slice,1:n  \_ lvcreate -y -L 20472s -n docker-poolmeta docker
=========================================================================

Udev at that particular moment doesn't process any messages. 

Here the output of udevadm monitor:
=========================================================================
# udevadm monitor 

KERNEL[3666.987063] remove   /devices/virtual/block/dm-7 (block)
KERNEL[3666.999446] remove   /kernel/slab/:t-0000320 (slab)
KERNEL[3667.000505] remove   /devices/virtual/bdi/253:7 (bdi)
KERNEL[3667.001290] remove   /devices/virtual/block/dm-7 (block)
KERNEL[3667.075290] add      /devices/virtual/net/docker0 (net)
KERNEL[3667.076606] add      /devices/virtual/net/docker0/queues/rx-0 (queues)
KERNEL[3667.076626] add      /devices/virtual/net/docker0/queues/tx-0 (queues)

========================================================================

There is no any UDEV processing message, just events from the kernel.

Further investigation shows, that udev was stopped

/var/log/puppet/fuelmenu-puppet.log
========================================================================
2016-07-13 07:57:32 +0000 Puppet (debug): Executing '/sbin/udevadm control --stop-exec-queue'
========================================================================

The reason is following peace of code, which stops udev execution for some reason:

https://github.com/openstack/fuel-library/blob/master/deployment/puppet/l23network/lib/puppet/provider/disable_hotplug/lnx.rb#L31

Conclusion:

The docker-storage hangs on semaphore, because to udev has been stopped for processing incoming events in queue. Udev is stopped by puppet during executing lnx.rb