Moby: Failed to connect to the Docker daemon

[wenchma]

Description of problem:

After install docker by running curl -sSL https://test.docker.com/ | sh , and DOCKER_HOST environment variable is not set in my shell, run docker cmd, but return :
$ docker ps
Cannot connect to the Docker daemon. Is the docker daemon running on this host?
$ service docker status
docker start/running, process 28966

docker version :

Client:
 Version:      1.9.0-rc5
 API version:  1.21
 Go version:   go1.4.2
 Git commit:   9318004
 Built:        Tue Nov  3 05:59:27 UTC 2015
 OS/Arch:      linux/amd64
Cannot connect to the Docker daemon. Is the docker daemon running on this host?

docker info :

Cannot connect to the Docker daemon. Is the docker daemon running on this host?

uname -a :

Linux docker 3.13.0-43-generic #72-Ubuntu SMP Mon Dec 8 19:35:06 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux

Environment details (AWS, VirtualBox, physical, etc.):OpenStack Cloud instance

Actual Results: docker daemon cannot work

Expected Results: docker daemon can work Ok

[Skarlso]

Could you please try running

sudo su -
service docker start
docker images

See if that works. And if it does, please add your user to the docker user group like this =>

sudo usermod -aG docker <username>

_Edit_: You might have to log out and than back in again for it to take effect.

[wenchma]

@Skarlso , I did "sudo usermod -aG docker ubuntu" , but not log out and back in to take effect.
Thanks, :smile:

[Skarlso]

Hah, I couldn't count on one hand how many times that happened to me as well. :D :+1:

[lordnynex]

To anyone who stumbles across this issue in the future.

I just spent a bit of time debugging this issue in a few of my servers. I now realize ssh ControlPersist multiplexing makes the whole 'log out log back in' thing not work the way you think it does. It makes sense why this happens, I just never thought about it happening like this before.

To fix this

$ ssh -O check -S ~/.ssh/[email protected] server.com
Master running (pid=14920)
$ ssh -O stop -S ~/.ssh/[email protected] server.com
Stop listening request sent.

Then log back into your server.

[kikiya]

I'm having the same problem but neither of the solutions provided worked.

service docker status

Redirecting to /bin/systemctl status  docker.service

● docker.service - Docker Application Container Engine
   Loaded: loaded (/usr/lib/systemd/system/docker.service; disabled; vendor preset: disabled)
   Active: active (running) since Wed 2015-12-16 18:19:56 EST; 17h ago
     Docs: https://docs.docker.com
 Main PID: 5181 (docker)
   Memory: 2.3M
   CGroup: /system.slice/docker.service
           └─5181 /usr/bin/docker daemon -H fd://

When I try to execute other commands, it shows "cannot connect to the docker daemon"

# docker version
Client:
 Version:      1.9.1
 API version:  1.21
 Go version:   go1.4.2
 Git commit:   a34a1d5
 Built:        Fri Nov 20 13:25:01 UTC 2015
 OS/Arch:      linux/amd64
Cannot connect to the Docker daemon. Is the docker daemon running on this host?

Has any one tried other troubleshooting? I'm running redhat/softlayer.

BTW, It was running hello-world fine, but now after following the above troubleshooting, it's showing the error when I try hello-world as well.

[kikiya]

Solved by manually starting the daemon. For some reason, the service is not actually starting it.

ran docker daemon & and now it's working fine. will have to investigate further to see why the service is not fully starting it.

[fflatorre]

We got exactly the same issue. Also for us "/usr/boin/docker -d" worked but we'd much prefer to be fixed when restarting via upstart.

[xecronix]

On Fedora 22

Installing Docker:
$> curl -fsSL https://get.docker.com/ | sh

After installing Docker:

A user needs to be added to the docker group.
$> sudo usermod -aG docker

The docker daemon needs to be started
$> sudo service docker start

You can set the daemon to start at boot
$> sudo chkconfig docker on

You can verify the docker service is running
$> service docker status

And one last final check
$> docker run hello-world

[ackris]

The issue is neatly documented in Docker Documentation.

My environment:
VirtualBox

OS: CentOS-7 -x86_64-Minimal

After adding user to the group, execute the following command.

sudo chkconfig docker on

At first, I couldn't view the server config details on the execution of docker version , and I received the same error "Cannot connect to the Docker daemon. Is the docker daemon running on this host?"

After executing sudo chkconfig docker on , everything works super fine :)

[hopeseekr]

On the latest Arch Linux, the only way I can get docker to work is to

sudo systemctl stop docker
sudo docker daemon &

Then it works appropriately.

After it is started by systemctl, I see this in the process list:

root 1933 0.0 0.1 360524 11748 ? Ssl 08:47 0:00 docker-containerd -l /var/run/docker/libcontainerd/docker-containerd.sock --runtime docker-runc

When I run it manually, I see:

root      2307  1.7  0.5 604268 34520 pts/0    Sl   08:56   0:00 docker daemon
root      2313  0.2  0.2 221256 12112 ?        Ssl  08:56   0:00 docker-containerd -l /var/run/docker/libcontainerd/docker-containerd.sock --runtime docker-runc

[thaJeztah]

@hopeseekr was that an upgrade from an existing installation, or a fresh install? Anything in the logs? Any modifications you made to the systemd unit file?

[hopeseekr]

Fresh install on a new Arch Linux install that I setup yesterday. I don't see any docker-specific log files in /var/log.

[hopeseekr]

I actually absolutely hate systemd, and I have NO IDEA how to debug or even look at what commands a systemd service actually runs. God! I miss OpenRC so much!

[thaJeztah]

@hopeseekr haha 🙊

you can probably get some information from sudo systemctl status docker and journalctl -xe -u docker. Did you use the package from our apt repositories or the arch-linux repos?

[hopeseekr]

Yes, I used the official packman -S docker route.

After a fresh reboot:

● docker.service - Docker Application Container Engine
   Loaded: loaded (/usr/lib/systemd/system/docker.service; enabled; vendor preset: disabl
   Active: active (running) since Tue 2016-05-24 09:28:34 CDT; 12s ago
     Docs: https://docs.docker.com
 Main PID: 7302 (docker)
    Tasks: 19 (limit: 512)
   Memory: 15.4M
      CPU: 134ms
   CGroup: /system.slice/docker.service
           ├─7302 /usr/bin/docker daemon -H fd://
           └─7308 docker-containerd -l /var/run/docker/libcontainerd/docker-containerd.so

May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.277017655-05:00" level=info msg="[graphdriver] using prior storage driver \"btrfs\""
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.289658653-05:00" level=info msg="Graph migration to content-addressability took 0.00 seconds"
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.295186622-05:00" level=info msg="Firewalld running: false"
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.341449086-05:00" level=info msg="Default bridge (docker0) is assigned with an IP address 172.17.0.0/16. Daemon option --bip can be used to set a preferred IP address"
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.388887470-05:00" level=info msg="Loading containers: start."
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.389074804-05:00" level=info msg="Loading containers: done."
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.389156085-05:00" level=info msg="Daemon has completed initialization"
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.389239746-05:00" level=info msg="Docker daemon" commit=5604cbe graphdriver=btrfs version=1.11.1
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.393658963-05:00" level=info msg="API listen on 0.0.0.0:2375"
May 24 09:28:34 libran systemd[1]: Started Docker Application Container Engine.

journalctl:

-- Subject: Unit docker.service has begun start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit docker.service has begun starting up.
May 24 09:28:33 libran docker[7302]: time="2016-05-24T09:28:33.271562128-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.277017655-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.289658653-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.295186622-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.341449086-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.388887470-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.389074804-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.389156085-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.389239746-05:00" level=inf
May 24 09:28:34 libran docker[7302]: time="2016-05-24T09:28:34.393658963-05:00" level=inf
May 24 09:28:34 libran systemd[1]: Started Docker Application Container Engine.
-- Subject: Unit docker.service has finished start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit docker.service has finished starting up.
-- 
-- The start-up result is done.

Netstat reports: tcp 0 0 0.0.0.0:2375 0.0.0.0:* LISTEN 1/init

But docker keeps saying:

Cannot connect to the Docker daemon. Is the docker daemon running on this host?

So then I run sudo killall -9 docker and then sudo docker daemon &, and the command shows:

WARN[0000] containerd: low RLIMIT_NOFILE changing to max  current=1024 max=4096
INFO[0001] [graphdriver] using prior storage driver "btrfs" 
INFO[0001] Graph migration to content-addressability took 0.00 seconds 
INFO[0001] Firewalld running: false                     
INFO[0001] Default bridge (docker0) is assigned with an IP address 172.17.0.0/16. Daemon option --bip can be used to set a preferred IP address 
INFO[0001] Loading containers: start.                   

INFO[0001] Loading containers: done.                    
INFO[0001] Daemon has completed initialization          
INFO[0001] Docker daemon                                 commit=5604cbe graphdriver=btrfs version=1.11.1
INFO[0001] API listen on /var/run/docker.sock           

and then it works correctly.

[tsmith@libran ~]$ docker images
REPOSITORY          TAG                 IMAGE ID            CREATED             SIZE
namespace/project   latest              74941504eebb        5 weeks ago         1.156 GB

[thaJeztah]

Yes, I used the official packman -S docker route.

Yes, those are packages maintained and supported by arch linux; it's possible that they ship with different options and configuration.

From the output you gave, it looks like there's something incorrectly set in the systemd unit-file, or there's a daemon.json present somewhere; the daemon that's started through systemd is listening on 0.0.0.0:2375;

API listen on 0.0.0.0:2375 

(note that that is a _dangerous_ configuration, if your host is accessible from the internet; it gives unprotected access to the API through port 2375 of your computer)

When you start it manually, it's using the defaults, and correctly listening on a local socket only;

INFO[0001] API listen on /var/run/docker.sock 

So;

• check if the systemd unit for docker has the right command-line options set for docker. you can see the unit-files we ship in the official packages here; https://github.com/docker/docker/tree/master/contrib/init/systemd
• it's possible there's a drop-in (override) file installed somewhere that overrides the default settings. see https://docs.docker.com/engine/admin/systemd/#custom-docker-daemon-options

[fayyazurrehman]

For ubuntu 14.04 on virtual box vm with docker installation

sudo groupadd docker
sudo gpasswd -a ${USER} docker
sudo service docker restart
Now, restart the OS. This solution is well explained here
https://jee-appy.blogspot.com/2016/02/installing-docker-in-virtual-box.html

[doublebyte1]

I experiment the same issue on Ubuntu Xenial (and on the previous version on ubuntu). Although the daemon is running, I cannot connect to it with my regular user (which is part of the docker group). Sometimes I restart, and it magically works! No idea why....

[jesseluo]

I got same issue on Cent OS 7 which I installed docker by yum. I tried all @xecronix mentioned here but still can't fix it.

At last I found I should run docker after sudo. The hint doesn't mean docker daemon isn't running...

[smelike]

I got the same issue in CentOS Linux release 7.2.1511 (Core) which I installed by "curl -sSL https://get.docker.com/ | sh". I tried "sudo docker daemon &", and the issue is fixed.

[doublebyte1]

In my case, if I check:
sudo service docker status
I can see the daemon is running, but I can still see the message: "_Cannot connect to the Docker daemon. Is the docker daemon running on this host?_" Running :
sudo service docker restart
Apparently solves the problem (I don't know why).

[NoviceF]

On ubuntu 14.04 for me workaround is not simply run docker (not service, but client) as sudo. Before run "docker run --rm busybox true", for example, move to sudo bash by typing "sudo su". In sudo bash docker is working fine.
Apparently, adding user to docker group will have a similar effect and in this case run bash as sudo is not necessary.

[LeoWinterDE]

fully restart the os, not just logging out and back in. was the solution for me.

[LaharlMontogmmery]

The use group changes doesn't apply immediately when a user is added in a group (the user need to log off and log on again for the changes to apply), but it is better to just restart the OS.

[Skarlso]

@dandv Done.

[dandv]

@wenchma: can you please correct the typo in the title of this issue, so searches can find it more easily? Also, "but not log out and back in to take effect." -> "but had to log out and back in."

[Skarlso]

No problem man. :-) glad I could help out folks. :-)

[jeffchiu2009]

On my latest Mac OSX Sierra,
I also got the error, after I install Docker Toolbox, when I run docker on my terminal:

git:(master) ✗ docker version
Client:
 Version:      1.12.2
 API version:  1.24
 Go version:   go1.6.3
 Git commit:   bb80604
 Built:        Tue Oct 11 17:00:50 2016
 OS/Arch:      darwin/amd64
Cannot connect to the Docker daemon. Is the docker daemon running on this host?

[johshoff]

Based on the number of people giving a thumbs up to @Skarlso's comment about

sudo usermod -aG docker <username>

maybe there should be check for that so you could say:

You need to be in the 'docker' group

instead of the misleading advice

Cannot connect to the Docker daemon. Is the docker daemon running on this host?

[thaJeztah]

@johshoff docker 1.13 will give a more informative error message if the client fails to connect due to permissions error; see this pull request; https://github.com/docker/docker/pull/26797

[Suncatcher]

I get on my AWS instance following stuff when I try to start Docker service

time="2017-02-01T12:34:46.999769865Z" level=info msg="libcontainerd: new containerd process, pid: 2737" 
time="2017-02-01T12:34:48.004600950Z" level=fatal msg="Error starting daemon: error initializing graphdriver: devicemapper: Non existing device docker-docker--pool" 

What should I do then?

P.S. Full log

[Skarlso]

@Suncatcher This Issues is closed. If you have a problem, please open a new one.

[DrST]

Dear all, just in case this would happen to solve you issue ...

I did all the abode, but only running docker daemon in interactive mode solved my problem.
When running the daemon like this, I got the missing information:

Error starting daemon: error initializing graphdriver: /var/lib/docker contains several valid graphdrivers: aufs, devicemapper; Please cleanup or explicitly choose storage driver (-s )

Previously I was using the old docker from Ubuntu 14.04 and upgraded it using the official upstream docker repository...

After specifying the correct devince mapper (aufs) a lot of checks were performed for existing images and everything was fine afterwards.

I hope this will help others :-)

[Suncatcher]

@DrST
What means _running docker daemon in interactive mode_?
Could you please describe the exact steps you did?

[westy92]

Should we fix the typo in the issue name? *Failed

[Suncatcher]

Should we fix the typo in the issue name? *Failed

Definitely!

[jpdonnell]

There underlying problem is the docker group isn't created ; so a usermod -G user docker will also fail

[Suncatcher]

Dunno where the problem is, but on standard Amazon linux Docker is installable like a charm, so I recommend to use standard.
But to ship ECS-optimized Linux AMI (yes-yes, pay attention to these letters - it's container-optimized) without ability to being configured for Docker is a bullshit! It's simply a shit!
*ucking Amazon is trying to make money from nowhere and it's frustrating.
Why the hell I have to be a DevOps-level to simply use Docker? They make me to buy their paid support plan, I see no other reason.

[soupault]

Had the same issue in AWS-hosted Archlinux.
Fixed via:

pacaur -S docker-machine
docker-machine env default
sudo systemctl start docker

[l-zhi]

centos 6.6 ,docker 1.7.1

> yum install device-mapper-event-libs
> sudo service docker restart

it's worked for me

[sironitomas]

I had the same problem on Arch Linux (Docker package version: 1:17.03.1-1). I modified this on /usr/lib/systemd/system/docker.service and I got it working fine after that.

#ExecStart=/usr/bin/dockerd -H fd://
ExecStart=/usr/bin/dockerd

[RakeshMsr]

Just follow below steps to resolve the issue.

$:sudo groupadd docker
$:sudo gpasswd -a ${USER} docker
$:sudo service docker restart
$:newgrp docker

[KingAndroid]

Thank you RakeshMsr.
You save my hard working and day.

[bronsonatcs]

any other fixes? none of this working for me.

[postarc]

Type:
$: dockerd
If you see a message: "Please upgrade your kernel to 3.10.0 or newer.", upgrade your linux kernel.