如何在centos 7的未配置Docker上删除/安装Docker映像

Question

Using Centos 6.6 and 7 and deciding to move to centos 7 as there are some issues using docker with centos 6.6 (reboot issues for me) and i'm trying to pull the current centos image from docker. (should just be docker pull centos )

However because i already had a docker image of centos installed on the 6.6 virtual machine, i thought it conflicts with the one im trying to pull on the centos 7. It states that the image (f1b something) is already being used on the system and is causing the download to not go through. Simply going over to the centos 6.6 and trying to remove the images (which would be labeled as none by the way, thus you have to do docker images -a ),even with force, does nothing. The only solution so far to this is to do a full removal of docker and its dependencies, and reinstall it which should come package free.

Of course this is not the solution i want. One of two things can happen. Either a way to make the two of them to coexist, or a way to remove the current one without removing any other current images. Or if I am not getting this right, take an entirely different approach.

EDIT+1: Ok heres the actual error im receiving when doing the the docker pull...

f1b...: download complete
f1b...:error downloading dependant layers
c85...:Downloading [>
7322...: Error pulling image (latest) from docker.io/centos, endpoint :https://registry-1.docker.io/v1,Dr
7322...:Error pulling image (latest) from docker.io/centos, Driver devicemapper failed to create image rootfs
FATA[0012] Error pulling image (latest) from docker.io/centos, Driver device mapper failed to create image rootfs f1b...:error running DeviceCreate (createSnapDevice) dm_task_run failed

And looking over the problem more im not so sure if its because of the centos 6.6 like i had initially thought despite sharing the same ID's.

EDIT +2: Stranger still is that the fatal error codes keeps changing (im assuming those are FATA[0012]?)

Answer 1

http://docker-sean.readthedocs.org/en/latest/chapter1.html

Theres a config file that needs to be changed for centos 7 docker users which happened to be applying the following change OPTIONS='-g /docker/data -p /var/run/docker.pid'

in the vim/vi file of /etc/sysconfig/docker .

I swear docker is going to be the death of me...

EDIT +1: Ok lets remap the solution to the following starting from a new centos 7 machine...

yum install docker
service docker start
docker pull centos

ERROR

systemctl enable docker.service

ERROR?

sudo systemctl enable docker.service
systemctl start docker.service

ERROR?

yum remove docker
yum install epel-release.noarch
yum install docker-io
vim /etc/sysconfig/docker
OPTIONS='-g /docker/data -p /var/run/docker.pid'
service docker restart
docker pull centos

and thats how i got docker to work on the new VM if i mapped it correctly. Also one of the commands i might have used was a thin_check. Somebody used it to verify docker in this link

EDIT +2:

Oh wow, this would explain even better whats happening here. See, the docker server can be installed straight out of the box with centos 7, however the daemon must still be installed from epel. As a reminder, the daemon is the item that actually runs the docker service. The server just allows docker to connect to the internet and view its repositories. Link is right here .