I want to know if what I'm doing is considered best practice, or is there a better way:
What I want
I want to have a Docker image, which will have some environment variables predefined when run as a container. I want to do that by running some shell script that will export those variables.
What I'm doing
My dockerfile looks like this:
Dockerfile code..
..
..
RUN useradd -m develop
RUN echo ". /env.sh" >> /home/develop/.bashrc
USER develop
Is that a good way?
Using the Dockerfile ENV directive would be much more usable than putting environment variables into a file.
ENV SOME_VARIABLE=some-value
# Do not use .bashrc at all
# Do not `RUN .` or `RUN source`
Most ways to use Docker don't involve running shell dotfiles like .bashrc. Adding settings there won't usually have any effect. In a Dockerfile, any environment variable settings in a RUN instruction will get lost at the end of that line, including files you read in using the shell . command (or the equivalent but non-standard source).
For example, given the Dockerfile you show, a docker run invocation like this never invokes a shell at all and never reads the .bashrc file:
docker run the-image env \
| grep A_VARIABLE_FROM_THE_BASHRC
There are some workarounds to this (my answer to How to source a script with environment variables in a docker build process? describes a startup-time entrypoint wrapper) but the two best ways are to (a) restructure your application to need fewer environment variables and have sensible defaults if they're not set, and (b) use ENV instead of a file of environment-variable settings.
** Check edits
Yes will probably work, if in the right order.
# Add user
RUN useradd -m develop
#switch to user
USER develop
#run script as user.
#RUN echo "./env.sh" >> /home/develop/.bashrc
RUN /bin/bash -c "source ./env.sh"
Although the duplicated RUN useradd is not necessary at all
I developed a few ROS packages and I want to put the packages in a docker container because installing all the ROS packages all the time is tedious. Therefore I created a dockerfile that uses a base ROS image, installed all the necessary dependencies, copied my workspace, built the workspace in the docker container and sourced everything afterward. You can find the docker file here:
FROM ros:kinetic-ros-base
RUN apt-get update && apt-get install locales
RUN locale-gen en_US.UTF-8
ENV LANG en_US.UTF-8
RUN apt-get update && apt-get install -y \
&& rm -rf /var/likb/apt/lists/*
COPY . /catkin_ws/src/
WORKDIR /catkin_ws
RUN /bin/bash -c '. /opt/ros/kinetic/setup.bash; catkin_make'
RUN /bin/bash -c '. /opt/ros/kinetic/setup.bash; source devel/setup.bash'
CMD ["roslaunch", "master_launch sim_perception.launch"]
The problem is: When I run the docker container wit the "run" command, docker doesn't seem to know that I sourced my new ROS workspace and therefore it cannot launch automatically my launch script. If I run the docker container as bash script with "run -it bash" I can source my workspace again and then roslaunch my .launch file.
So can someone tell me how to write my dockerfile correctly so I launch my .launch file automatically when I run the container? Thanks!
From Docker Docs
Each RUN instruction is run independently and won't effect next instruction so when you run last Line no PATH are saved from ROS.
You need Source .bashrc or every environment you need using source first.
You can wrap everything you want (source command and roslaunch command) inside a sh file then just run that file at the end
If you review the convention of ros_entrypoint.sh you can see how best to source the workspace you would like in the docker. We're all so busy learning how to make docker and ros do the real things, it's easy to skip over some of the nuance of this interplay. This sucked forever for me; hope this is helpful for you.
I looked forever and found what seemed like only bad advice, and in the absence of an explicit standard or clear guidance I've settled into what seems like a sane approach that also allows you to control what launches at runtime with environment variables. I now consider this as the right solution for my needs.
In the Dockerfile for the image you want to set the start/launch behavior;
towards the end; you should use ADD line to insert your own ros_entrypoint.sh (example included); Set it as the ENTRYPOINT and then a CMD to run by default run something when the docker start.
note: you'll (obviously?) need to run the docker build process for these changes to be effective
Dockerfile looks like this:
all your other dockerfile ^^
.....
# towards the end
COPY ./ros_entrypoint.sh /
ENTRYPOINT ["/ros_entrypoint.sh"]
CMD ["bash"]
Example ros_entryppoint.sh:
#!/bin/bash
set -e
# setup ros environment
if [ -z "${SETUP}" ]; then
# basic ros environment
source "/opt/ros/$ROS_DISTRO/setup.bash"
else
#from environment variable; should be a absolute path to the appropriate workspaces's setup.bash
source $SETUP
fi
exec "$#"
Used in this way the docker will automatically source either the basic ros bits... or if you provide another workspace's setup.bash path in the $SETUP environment variable, it will be used in the container.
So a few ways to work with this:
From the command line prior to running docker
export SETUP=/absolute/path/to/the/setup.bash
docker run -it your-docker-image
From the command line (inline)
docker run --env SETUP=/absolute/path/to/the/setup.bash your-docker-image
From docker-compose
service-name:
network_mode: host
environment:
- SETUP=/absolute/path/to/the_workspace/devel/setup.bash #or whatever
command: roslaunch package_name launchfile_that_needed_to_be_sourced.launch
#command: /bin/bash # wake up and do something else
Say I have a Dockerfile:
.
.
RUN echo 'source /root/script.sh' >> /etc/bash.bashrc
(The script adds some env variables)
If I:
1) Do this:
docker run -it -v /home/user/script.sh:/root/script.sh image
It takes me to shell where if I call "env" I see the variable set by the script
But if I:
2) Do this:
docker run -it -v /home/user/script.sh:/root/script.sh image env
It prints out env and exits and my variable is missing
What am I missing? I need the variable to exists even if I specify a command/script like "env" at the end of the docker run command
When you run a command like
docker run ... image command
Docker directly runs the command you give; it doesn’t launch any kind of shell, and there’s no opportunity for a .bashrc or similar file to be read.
I’d suggest two things here:
If your program does need environment variables set in some form, set them directly using Dockerfile ENV directives. Don’t try to edit .bashrc or /etc/profile or any other shell dotfile; they won’t reliably get run.
As much as you can install things in places so that you don’t need to change environment variables. For instance, Python supports a “virtual environment” concept that allows an isolated library environment, which requires changing $PATH and similar things; but Docker provides the same isolation on its own, so just install things into the “global” package space.
If you really can’t manage either of these things, then you can write an entrypoint script that sets environment variables and then launches the container’s command. This might look like
#!/bin/sh
. /root/script.sh
exec "$#"
And then you could include this in your Dockerfile like
...
COPY entrypoint.sh /
ENTRYPOINT ["/entrypoint.sh"]
CMD ["/app/myapp"]
(If you need to use docker exec to get a debugging shell in the container, that won’t be a child process of the entrypoint and won’t get its environment variables.)
When I try to use an environment variable($HOME) that I set in the Dockerfile, in the script that runs at start up, $HOME is not set. If I run printenv in the container, $HOME is set. So I am confused, and not sure what is going on.
I am using the phusion/passenger-customizable image, so that I can run a custom node server via pm2. I need a different version of Node then what is bundled in the node specific passenger image.
Dockerfile
# Simplified
FROM phusion/passenger-customizable:0.9.27
RUN apt-get update && apt-get upgrade -y -o Dpkg::Options::="--force-confold"
# Set environment variables needed for the docker image.
ARG HOME=/opt/var/app
ENV HOME $HOME
# Use baseimage-docker's init process.
CMD ["/sbin/my_init"]
RUN mkdir /etc/service/app
ADD start.sh /etc/service/app/run
RUN chmod a+x /etc/service/app/run
start.sh
echo $HOME
# run some scripts that reference the $HOME directory
What do I need to do to be able to reference a environment variable, set in the Dockerfile, in my start up scripts? Or do I just need to hardcode the paths in that start up script and call it a day?
$HOME is reserved, in some fashion. When running printenv, per #Sebastian, all my other variables where there but not $HOME. I prepended it with the initials of my company and it is working as intended.
I have a docker image which installs grunt, but when I try to run it, I get an error:
Error response from daemon: Cannot start container foo_1: \
exec: "grunt serve": executable file not found in $PATH
If I run bash in interactive mode, grunt is available.
What am I doing wrong?
Here is my Dockerfile:
# https://registry.hub.docker.com/u/dockerfile/nodejs/ (builds on ubuntu:14.04)
FROM dockerfile/nodejs
MAINTAINER My Name, me#email.com
ENV HOME /home/web
WORKDIR /home/web/site
RUN useradd web -d /home/web -s /bin/bash -m
RUN npm install -g grunt-cli
RUN npm install -g bower
RUN chown -R web:web /home/web
USER web
RUN git clone https://github.com/repo/site /home/web/site
RUN npm install
RUN bower install --config.interactive=false --allow-root
ENV NODE_ENV development
# Port 9000 for server
# Port 35729 for livereload
EXPOSE 9000 35729
CMD ["grunt"]
This was the first result on google when I pasted my error message, and it's because my arguments were out of order.
The container name has to be after all of the arguments.
Bad:
docker run <container_name> -v $(pwd):/src -it
Good:
docker run -v $(pwd):/src -it <container_name>
When you use the exec format for a command (e.g., CMD ["grunt"], a JSON array with double quotes), it will be executed without a shell. This means that most environment variables will not be present.
If you specify your command as a regular string (e.g. CMD grunt) then the string after CMD will be executed with /bin/sh -c.
More info on this is available in the CMD section of the Dockerfile reference.
I found the same problem. I did the following:
docker run -ti devops -v /tmp:/tmp /bin/bash
When I change it to
docker run -ti -v /tmp:/tmp devops /bin/bash
it works fine.
For some reason, I get that error unless I add the "bash" clarifier. Even adding "#!/bin/bash" to the top of my entrypoint file didn't help.
ENTRYPOINT [ "bash", "entrypoint.sh" ]
There are several possible reasons for an error like this.
In my case, it was due to the executable file (docker-entrypoint.sh from the Ghost blog Dockerfile) lacking the executable file mode after I'd downloaded it.
Solution: chmod +x docker-entrypoint.sh
I had the same problem, After lots of googling, I couldn't find out how to fix it.
Suddenly I noticed my stupid mistake :)
As mentioned in the docs, the last part of docker run is the command you want to run and its arguments after loading up the container.
NOT THE CONTAINER NAME !!!
That was my embarrassing mistake.
Below I provided you with the picture of my command line to see what I have done wrong.
And this is the fix as mentioned in the docs.
A Docker container might be built without a shell (e.g. https://github.com/fluent/fluent-bit-docker-image/issues/19).
In this case, you can copy-in a statically compiled shell and execute it, e.g.
docker create --name temp-busybox busybox:1.31.0
docker cp temp-busybox:/bin/busybox busybox
docker cp busybox mycontainerid:/busybox
docker exec -it mycontainerid /bin/busybox sh
In the error message shown:
Error response from daemon: Cannot start container foo_1: \
exec: "grunt serve": executable file not found in $PATH
It is complaining that it cannot find the executable grunt serve, not that it could not find the executable grunt with the argument serve. The most likely explanation for that specific error is running the command with the json syntax:
[ "grunt serve" ]
in something like your compose file. That's invalid since the json syntax requires you to split up each parameter that would normally be split by the shell on each space for you. E.g.:
[ "grunt", "serve" ]
The other possible way you can get both of those into a single parameter is if you were to quote them into a single arg in your docker run command, e.g.
docker run your_image_name "grunt serve"
and in that case, you need to remove the quotes so it gets passed as separate args to the run command:
docker run your_image_name grunt serve
For others seeing this, the executable file not found means that Linux does not see the binary you are trying to run inside your container with the default $PATH value. That could mean lots of possible causes, here are a few:
Did you remember to include the binary inside your image? If you run a multi-stage image, make sure that binary install is run in the final stage. Run your image with an interactive shell and verify it exists:
docker run -it --rm your_image_name /bin/sh
Your path when shelling into the container may be modified for the interactive shell, particularly if you use bash, so you may need to specify the full path to the binary inside the container, or you may need to update the path in your Dockerfile with:
ENV PATH=$PATH:/custom/dir/bin
The binary may not have execute bits set on it, so you may need to make it executable. Do that with chmod:
RUN chmod 755 /custom/dir/bin/executable
The binary may include dynamically linked libraries that do not exist inside the image. You can use ldd to see the list of dynamically linked libraries. A common reason for this is compiling with glibc (most Linux environments) and running with musl (provided by Alpine):
ldd /path/to/executable
If you run the image with a volume, that volume can overlay the directory where the executable exists in your image. Volumes do not merge with the image, they get mounted in the filesystem tree same as any other Linux filesystem mount. That means files from the parent filesystem at the mount point are no longer visible. (Note that named volumes are initialized by docker from the image content, but this only happens when the named volume is empty.) So the fix is to not mount volumes on top of paths where you have executables you want to run from the image.
If you run a binary for a different platform, and haven't configured binfmt_misc with the --fix-binary option, qemu will be looking for the interpreter inside the container filesystem namespace instead of the host filesystem. See this Ubuntu bug report for more details on this issue.
If the error is from a shell script, the issue is often with the first line of that script (e.g. the #!/bin/bash). Either the command doesn't exist inside the image for a reason above, or the file is not saved as ascii or utf8 with Linux linefeeds. You can attempt dos2unix to fix the linefeeds, or check your git and editor settings.
in my case i order params wrong move all switchs before image name
I got this error message, when I was building alpine base image :
ERROR: for web Cannot start service web: OCI runtime create failed: container_linux.go:380: starting container process caused: exec: "bash": executable file not found in $PATH: unknown
In my docker-compose file, I had the command directive in which executing command using bash and bash does not come with alpine base image.
command: bash -c "python manage.py runserver 0.0.0.0:8000"
Then I realized and executed command using sh (shell).
It worked for me.
problem is glibc, which is not part of apline base iamge.
After adding it worked for me :)
Here are the steps to get the glibc
apk --no-cache add ca-certificates wget
wget -q -O /etc/apk/keys/sgerrand.rsa.pub https://alpine-pkgs.sgerrand.com/sgerrand.rsa.pub
wget https://github.com/sgerrand/alpine-pkg-glibc/releases/download/2.28-r0/glibc-2.28-r0.apk
apk add glibc-2.28-r0.apk
Refering to the title.
My mistake was to put variables via --env-file during docker run. Among others the file consisted of a PATH extension: PATH=$PATH:something, which caused PATH var look literally like PATH=$PATH:something (var resolution hadn't been performed) instead of PATH:/usr/bin...:something.
I couldn't make the resolution work through --env-file, so the only way I see this working is by using ENV in Dockerfile.
I ran into this issue using docker-compose. None of the solutions here or on this related question resolved my issue. Ultimately what worked for me was clearing all cached docker artifacts with docker prune -a and restarting docker.
to make it work add soft reference to /usr/bin:
ln -s $(which node) /usr/bin/node
ln -s $(which npm) /usr/bin/npm