I'm getting started working with Docker. I'm using the WordPress base image and docker-compose.
I'm trying to ssh into one of the containers to inspect the files/directories that were created during the initial build. I tried to run docker-compose run containername ls -la, but that didn't do anything. Even if it did, I'd rather have a console where I can traverse the directory structure, rather than run a single command. What is the right way to do this with Docker?
docker attach will let you connect to your Docker container, but this isn't really the same thing as ssh. If your container is running a webserver, for example, docker attach will probably connect you to the stdout of the web server process. It won't necessarily give you a shell.
The docker exec command is probably what you are looking for; this will let you run arbitrary commands inside an existing container. For example:
docker exec -it <mycontainer> bash
Of course, whatever command you are running must exist in the container filesystem.
In the above command <mycontainer> is the name or ID of the target container. It doesn't matter whether or not you're using docker compose; just run docker ps and use either the ID (a hexadecimal string displayed in the first column) or the name (displayed in the final column). E.g., given:
$ docker ps
d2d4a89aaee9 larsks/mini-httpd "mini_httpd -d /cont 7 days ago Up 7 days web
I can run:
$ docker exec -it web ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
18: eth0: <BROADCAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
link/ether 02:42:ac:11:00:03 brd ff:ff:ff:ff:ff:ff
inet 172.17.0.3/16 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::42:acff:fe11:3/64 scope link
valid_lft forever preferred_lft forever
I could accomplish the same thing by running:
$ docker exec -it d2d4a89aaee9 ip addr
Similarly, I could start a shell in the container;
$ docker exec -it web sh
/ # echo This is inside the container.
This is inside the container.
/ # exit
$
To bash into a running container, type this:
docker exec -t -i container_name /bin/bash
or
docker exec -ti container_name /bin/bash
or
docker exec -ti container_name sh
Historical note: At the time I wrote this answer, the title of the question was: "How to ssh into a docker container?"
As other answers have demonstrated, it is common to execute and interact with preinstalled commands (including shells) in a locally-accessible running container using docker exec, rather than SSH:
docker exec -it (container) (command)
Note: The below answer is based on Ubuntu (of 2016). Some translation of the installation process will be required for non-Debian containers.
Let's say, for reasons that are your own, you really do want to use SSH. It takes a few steps, but it can be done. Here are the commands that you would run inside the container to set it up...
apt-get update
apt-get install openssh-server
mkdir /var/run/sshd
chmod 0755 /var/run/sshd
/usr/sbin/sshd
useradd --create-home --shell /bin/bash --groups sudo username ## includes 'sudo'
passwd username ## Enter a password
apt-get install x11-apps ## X11 demo applications (optional)
ifconfig | awk '/inet addr/{print substr($2,6)}' ## Display IP address (optional)
Now you can even run graphical applications (if they are installed in the container) using X11 forwarding to the SSH client:
ssh -X username#IPADDRESS
xeyes ## run an X11 demo app in the client
Here are some related resources:
openssh-server doesn't start in Docker container
How to get bash or ssh into a running container in background mode?
Can you run GUI applications in a Linux Docker container?
Other useful approaches for graphical access found with search: Docker X11
If you run SSHD in your Docker containers, you're doing it wrong!
If you're here looking for a Docker Compose-specific answer like I was, it provides an easy way in without having to look up the generated container ID.
docker-compose exec takes the name of the service as per your docker-compose.yml file.
So to get a Bash shell for your 'web' service, you can do:
$ docker-compose exec web bash
If the container has already exited (maybe due to some error), you can do
$ docker run --rm -it --entrypoint /bin/ash image_name
or
$ docker run --rm -it --entrypoint /bin/sh image_name
or
$ docker run --rm -it --entrypoint /bin/bash image_name
to create a new container and get a shell into it. Since you specified --rm, the container would be deleted when you exit the shell.
Notice: this answer promotes a tool I've written.
I've created a containerized SSH server that you can 'stick' to any running container. This way you can create compositions with every container. The only requirement is that the container has Bash.
The following example would start an SSH server attached to a container with name 'my-container'.
docker run -d -p 2222:22 \
-v /var/run/docker.sock:/var/run/docker.sock \
-e CONTAINER=my-container -e AUTH_MECHANISM=noAuth \
jeroenpeeters/docker-ssh
ssh localhost -p 2222
When you connect to this SSH service (with your SSH client of choice) a Bash session will be started in the container with name 'my-container'.
For more pointers and documentation see: https://github.com/jeroenpeeters/docker-ssh
Start a session into a Docker container using this command:
sudo docker exec -i -t (container ID) bash
If you're using Docker on Windows and want to get shell access to a container, use this:
winpty docker exec -it <container_id> sh
Most likely, you already have Git Bash installed. If you don't, make sure to install it.
In some cases your image can be Alpine-based. In this case it will throw:
OCI runtime exec failed: exec failed: container_linux.go:348: starting
container process caused "exec: \"bash\": executable file not found in
$PATH": unknown
Because /bin/bash doesn't exist. Instead of this you should use:
docker exec -it 9f7d99aa6625 ash
or
docker exec -it 9f7d99aa6625 sh
To connect to cmd in a Windows container, use
docker exec -it d8c25fde2769 cmd
Where d8c25fde2769 is the container id.
docker exec -it <container_id or name> bash
OR
docker exec -it <container_id or name> /bin/bash
GOINSIDE SOLUTION
install goinside command line tool with:
sudo npm install -g goinside
and go inside a docker container with a proper terminal size with:
goinside docker_container_name
old answer
We've put this snippet in ~/.profile:
goinside(){
docker exec -it $1 bash -c "stty cols $COLUMNS rows $LINES && bash";
}
export -f goinside
Not only does this make everyone able to get inside a running container with:
goinside containername
It also solves a long lived problem about fixed Docker container terminal sizes. Which is very annoying if you face it.
Also if you follow the link you'll have command completion for your docker container names too.
To inspect files, run docker run -it <image> /bin/sh to get an interactive terminal. The list of images can be obtained by docker images. In contrary to docker exec this solution works also in case when an image doesn't start (or quits immediately after running).
It is simple!
List out all your Docker images:
sudo docker images
On my system it showed the following output:
REPOSITORY TAG IMAGE ID CREATED VIRTUAL SIZE
bash latest 922b9cc3ea5e 9 hours ago
14.03 MB
ubuntu latest 7feff7652c69 5 weeks ago 81.15 MB
I have two Docker images on my PC. Let's say I want to run the first one.
sudo docker run -i -t ubuntu:latest /bin/bash
This will give you terminal control of the container. Now you can do all type of shell operations inside the container. Like doing ls will output all folders in the root of the file system.
bin boot dev etc home lib lib64 media mnt opt proc root run sbin srv sys tmp usr var
Simple
docker exec -it <container_id> bash
in above -it means interactive terminal.
also, with image name:
docker exec -it <REPOSITORY name> bash
I've created a terminal function for easier access to the container's terminal. Maybe it's useful to you guys as well:
So the result is, instead of typing:
docker exec -it [container_id] /bin/bash
you'll write:
dbash [container_id]
Put the following in your ~/.bash_profile (or whatever else that works for you), then open a new terminal window and enjoy the shortcut:
#usage: dbash [container_id]
dbash() {
docker exec -it "$1" /bin/bash
}
2022 Solution
Consider another option
Why do you need it?
There is a bunch of modern docker-images that are based on distroless base images (they don't have /bin/bash either /bin/sh) so it becomes impossible to docker exec -it {container-name} bash into them.
How to shell-in any container
Use opener:
requires to add alias in your environment opener wordpress
works anywhere docker run --rm -it -v /var/run/docker.sock:/var/run/docker.sock artemkaxboy/opener wordpress
Instead of wordpress you can use name or id or image-name of any container you want to connect
How it works
Opener is a set of python scripts wrapped-up to a docker image. It finds target container by any unique attribute (name, id, port, image), tries to connect to target using bash. If bash is not found opener tries to connect using sh. Finally if sh is not found either opener installs busybox into target container and connects to the target using busybox shell, opener deletes busybox during disconnection.
$ docker exec -it <Container-Id> /bin/bash
Or depending on the shell, it can be
$ docker exec -it <Container-Id> /bin/sh
You can get the container-Id via docker ps command
-i = interactive
-t = to allocate a psuedo-TTY
you can interact with the terminal in docker container by passing the option -ti
docker run --rm -ti <image-name>
eg: docker run --rm -ti ubuntu
-t stands for terminal
-i stands for interactive
There are at least 2 options depending on the target.
Option 1: Create a new bash process and join into it (easier)
Sample start: docker exec -it <containername> /bin/bash
Quit: type exit
Pro: Does work on all containers (not depending on CMD/Entrypoint)
Contra: Creates a new process with own session and own environment-vars
Option 2: Attach to the already running bash (better)
Sample start: docker attach --detach-keys ctrl-d <containername>
Quit: use keys ctrl and d
Pro: Joins the exact same running bash which is in the container. You have same the session and same environment-vars.
Contra: Only works if CMD/Entrypoint is an interactive bash like CMD ["/bin/bash"] or CMD ["/bin/bash", "--init-file", "myfile.sh"] AND if container has been started with interactive options like docker run -itd <image> (-i=interactive, -t=tty and -d=deamon [opt])
We found option 2 more useful. For example we changed apache2-foreground to a normal background apache2 and started a bash after that.
docker exec will definitely be a solution. An easy way to work with the question you asked is by mounting the directory inside Docker to the local system's directory.
So that you can view the changes in local path instantly.
docker run -v /Users/<path>:/<container path>
Use:
docker attach <container name/id here>
The other way, albeit there is a danger to it, is to use attach, but if you Ctrl + C to exit the session, you will also stop the container. If you just want to see what is happening, use docker logs -f.
:~$ docker attach --help
Usage: docker attach [OPTIONS] CONTAINER
Attach to a running container
Options:
--detach-keys string Override the key sequence for detaching a container
--help Print usage
--no-stdin Do not attach STDIN
--sig-proxy Proxy all received signals to the process (default true)
Use this command:
docker exec -it containerid /bin/bash
To exec into a running container named test, below is the following commands
If the container has bash shell
docker exec -it test /bin/bash
If the container has bourne shell and most of the cases it's present
docker run -it test /bin/sh
If you have Docker installed with Kitematic, you can use the GUI. Open Kitematic from the Docker icon and in the Kitematic window select your container, and then click on the exec icon.
You can see the container log and lots of container information (in settings tab) in this GUI too.
In my case, for some reason(s) I need to check all the network involved information in each container. So the following commands must be valid in a container...
ip
route
netstat
ps
...
I checked through all these answers, none were helpful for me. I’ve searched information in other websites. I won’t add a super link here, since it’s not written in English. So I just put up this post with a summary solution for people who have the same requirements as me.
Say you have one running container named light-test. Follow the steps below.
docker inspect light-test -f {{.NetworkSettings.SandboxKey}}. This command will get reply like /var/run/docker/netns/xxxx.
Then ln -s /var/run/docker/netns/xxxx /var/run/netns/xxxx. The directory may not exist, do mkdir /var/run/netns first.
Now you may execute ip netns exec xxxx ip addr show to explore network world in container.
PS. xxxx is always the same value received from the first command. And of course, any other commands are valid, i.e. ip netns exec xxxx netstat -antp|grep 8080.
There are two options we can connect to the docker terminal directly with these method shell and bash but usually bash is not supported and defualt sh is supported terminal
To sh into the running container, type this:
docker exec -it container_name/container_ID sh
To bash into a running container, type this:
docker exec -it container_name/container_ID bash
and you want to use only bash terminal than you can install the bash terminal in your Dockerfile like RUN apt install bash -y
This is best if you don't want to specify an entry point in your docker build file..
sudo docker run -it --entrypoint /bin/bash <container_name>
If you are using Docker Compose then this will take you inside a Docker container.
docker-compose run container_name /bin/bash
Inside the container it will take you to WORKDIR defined in the Dockerfile. You can change your work directory by
WORKDIR directory_path # E.g /usr/src -> container's path
Another option is to use nsenter.
PID=$(docker inspect --format {{.State.Pid}} <container_name_or_ID>)
nsenter --target $PID --mount --uts --ipc --net --pid
"sudo systemctl enable --now docker" while running this command I'm getting an error like "System has not been booted with systemd as init system (PID 1). Can't operate.
Failed to connect to bus: Host is down"
How can I fix this and I'm new to Ubuntu commands.
Please refer to Latest Docker Documentation; Older versions of Docker don't play nicely.
The trick is to uninstall sudo apt-get remove docker docker.io containderd, etc... And add the docker GPG key and setup the repository (documented in link). Then install sudo apt-get install docker-ce docker-ce-cli containerd.io instead.
Once you've removed old docker from your Ubuntu installation, you can then go sudo service docker start to have docker daemon running without systemctl and use service instead.
I am using ubuntu:20.04 Image.
I encountered the same error while trying to use systemctl command inside the ubuntu docker container, since the system has not been booted with systemd as init system (PID 1) as the error says so in order to fix that you will need to start you container using this
docker run -ti -d ubuntu:20.04 "/sbin/init"
In fact, running the container will cause an error:
I have solvedcontroller_1 | Failed to mount tmpfs at /run: Operation not permitted
container_1 | Failed to mount tmpfs at /run/lock: Operation not permitted
container_1 | [!!!!!!] Failed to mount API filesystems.
container_1 | Freezing execution.
By default, the container is not authorized to access any device, but a container with a "privileged" tag grants its root capabilities to all devices on the host system that will give access to the container to mount API filesystems and solve the issue.
in order to fix that you will need to
docker run -ti -d --privileged ubuntu:20.04 "/sbin/init"
I have a host with ubuntu 20.04, and I run firefox in container from ubuntu:20.04 image.
When firefox is already started on the host: container stops immediately, new window of firefox appears, and I can see all my host browsing history, sessions and so on.
When firefox is NOT started on the host: container is running, new window of "firefox [container hash]" appears, I can see only container browsing history and sessions there (as expected). BUT when I start firefox on the host while container is still running: new window of "firefox [same container hash]" appears, and I can see only container browsing history and sessions.
If I run firefox as a different user, like
sudo -H -u some-user firefox
and having umask 077 - I've got perfect isolation and parallel running without docker, but that's not the full goal
My dockerfile:
FROM ubuntu:20.04
WORKDIR /usr/src/app
RUN apt-get update && apt-get install -y firefox
CMD firefox
Terminal history:
xhost +local:docker
docker build -t firefox .
docker create -ti -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix --name ff firefox
docker start ff
I suppose this behavior of process launch from container is not really obvious and expected. Could you please explain what exactly is happening and why?
Docker container is not an isolated machine. The commands that run inside docker container are executed on the host machine (or the docker VM if using Docker for Mac).
This can be verified in the following way:
Run a command inside docker container docker exec -it <container-name> sleep 100
On the host machine, grep for this command ps -ef | grep sleep. For mac, docker run -it --privileged --pid=host debian nsenter -t 1 -m -u -n -i sh will provide a shell into the running docker VM.
On my machine:
# ps -ef | grep sleep
2609 root 0:00 sleep 100
2616 root 0:00 grep sleep
When you run a daemon, it creates a socket file in temp directory.
This file is the gateway to communication with the application.
For instance, when mysql is running in the system, it creates a socket file /var/run/mysqld/mysqld.sock which is used for communication by mysql client.
These daemons can also bind to a port, and be accessed through the network this way. These ports are simply socket connections to your application which are visible over the network.
Coming back to your question,
docker create -ti -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix --name ff firefox
/tmp/.X11-unix is managing Unix-domain sockets. Since this is mounted within the container, the socket space between the container and host is shared.
When firefox is running on the host, the socket is occupied already. Thus the container fails to start
When firefox is not running on host and container is started, the socket is free and hence the container is able to start. This uses the filesystem inside container to store history etc. Thus you do not see the history from host.
If you run firefox from host now, it will simply connect to this unix socket and launch a firefox window.
I have following scenario.
Two Machine ( Physical Machine)
One is Windows 10 With Docker On Windows Installer and same way ubuntu 18.04 with docker-ce installed.
I can run command on individual and that is fine.
I want to connect Ubuntu Docker Host from Docker on Windows machine. So Docker CLI on Windows Point to deamon at Ubuntu Host.
You will need to enable docker remote API on Ubuntu Docker Host by adding below settings in daemon.json or your startup script
[root#localhost ~]# cat /etc/docker/daemon.json
{
"hosts": [ "unix:///var/run/docker.sock", "tcp://0.0.0.0:2376" ]
}
Once you restart docker you can connect to docker host locally by socket file and remotely by listening port (2376).
Verify the listening port of docker on Ubuntu
[root#localhost ~]# netstat -ntlp | grep 2376
tcp6 0 0 :::2376 :::* LISTEN 1169/dockerd
Now you can connect to this docker from Windows machine by setting the DOCKER_HOST env variable in Windows like this
PS C:\Users\YellowDog> set DOCKER_HOST=tcp://<Ubuntu-Docker_Host-IP>:2376
PS C:\Users\YellowDog> docker ps
It will list docker containers running on Ubuntu Docker Host
You can also do this through additional options to the service:
Find original ExecStart line docker.service:
systemctl status docker | grep load | grep -oP "\/.+service"
# --> /lib/systemd/system/docker.service
cat /lib/systemd/system/docker.service | grep ExecStart
ExecStart=/usr/bin/dockerd -H fd:// $DOCKER_OPTS
Create a new file to store the daemon options:
sudo mkdir /etc/systemd/system/docker.service.d/
Add next lines with -H unix:// -H tcp://0.0.0.0:2375 options to /etc/systemd/system/docker.service.d/options.conf:
cat <<EOF > /etc/systemd/system/docker.service.d/options.conf
[Service]
ExecStart=
ExecStart=/usr/bin/dockerd -H fd:// \$DOCKER_OPTS -H unix:// -H tcp://0.0.0.0:2375
EOF
Here you need to pay attention to the escaping $DOCKER_OPTS variable if it exists.
Or using your favorite editor, for example vim.
Now, reload the systemd daemon and restart the docker service:
# Reload the systemd daemon.
sudo systemctl daemon-reload
# Restart Docker.
sudo systemctl restart docker
Configuring your dev box to connect to the remote Docker daemon:
If you want to set DOCKER_HOST by default so it always connects remotely you can export it in your ~/.bashrc file.
Here’s an example of that as a 1 liner:
echo "export DOCKER_HOST=tcp://X.X.X.X:2375" >> ~/.bashrc && source ~/.bashrc
Or use it all at once:
DOCKER_HOST=tcp://X.X.X.X:2375 docker ps
I would like to execute netstat inside a running docker container to see open TCP sockets and their statuses. But, on some of my docker containers, netstat is not available. Is there any way to get open sockets (and their statuses, and which IP addresses they are connected to if any) without using netstat, via some docker API? (BTW, my container uses docker-proxy - that is, not directly bridged)
I guess I could look at /proc file system directly, but at that point, I might as well docker cp netstat into the container and execute it. I was wondering if there was any facility that docker might provide for this.
You can use the nsenter command to run a command on your host inside the network namespace of the Docker container. Just get the PID of your Docker container:
docker inspect -f '{{.State.Pid}}' container_name_or_id
For example, on my system:
$ docker inspect -f '{{.State.Pid}}' c70b53d98466
15652
And once you have the PID, use that as the argument to the target (-t) option of nsenter. For example, to run netstat inside the container network namespace:
$ sudo nsenter -t 15652 -n netstat
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State
tcp 0 0 0.0.0.0:80 0.0.0.0:* LISTEN
Notice that this worked even though the container does not have netstat installed:
$ docker exec -it c70b53d98466 netstat
rpc error: code = 13 desc = invalid header field value "oci runtime error: exec failed: container_linux.go:247: starting container process caused \"exec: \\\"netstat\\\": executable file not found in $PATH\"\n"
(nsenter is part of the util-linux package)
The two commands from #larsks answer merged into one-liner - no need to copy-paste the PID(s) (just replace container_name_or_id):
sudo nsenter -t $(docker inspect -f '{{.State.Pid}}' container_name_or_id) -n netstat
If you have iproute2 package installed, you can use
sudo nsenter -t $(docker inspect -f '{{.State.Pid}}' container_name_or_id) -n ss
or
sudo nsenter -t $(docker inspect -f '{{.State.Pid}}' container_name_or_id) -n ss -ltu
It will show TCP and UDP
If you want them all (all containers) try this.
$ for i in `docker ps -q` ; do sudo nsenter -t $(docker inspect -f '{{.State.Pid}}' $i) -n netstat ; done
I tried the other solutions and it didn't work for me by my colleague gave me this solution. Thought I would mention it here for others like me and for me to refer to later lol.
docker exec -it [container name] bash
grep -v “rem_address” /proc/net/tcp
docker inspect <container_id>
Look for "ExposedPorts" in "Config"
server:docker container ls
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
80acfa804b59 admirito/gsad:10 "docker-entrypoint.s…" 18 minutes ago Up 10 minutes 80/tcp gvmcontainers_gsad_1