I am trying to integrate docker into my CI platform. After getting this working properly with a Docker-in-a-docker solution, I came across a blog post by one of the Docker maintainers, where he says that instead of using a Docker-in-a-docker solution for my CI, I should instead simply mount the /var/run/docker.sock to my CI container.
https://jpetazzo.github.io/2015/09/03/do-not-use-docker-in-docker-for-ci/
Simply put, when you start your CI container (Jenkins or other), instead of hacking something together with Docker-in-Docker, start it with:
docker run -v /var/run/docker.sock:/var/run/docker.sock ...
So I tried this. I ran the following command:
docker run -p 8080:8080 -p 50000:50000 -v /var/run/docker.sock:/var/run/docker.sock jenkins
Using jenkins as my CI container.
When running the above command, jenkins starts up properly, and I can jump into the container to see that the docker.sock file is located in the /var/run/ path.
However, when I run the command: docker, the machine returns with the following message:
bash: docker: command not found
Does anyone know what I am missing in order to make this work per the author's instructions?
I am using Docker v. 1.11.1, on a fresh CentOS 7 box.
Thanks in advance
Figured this out today. The above command will work so long as the docker daemon + dependencies are added to the container. In my case, I ended up writing a simple Dockerfile, which also included the line:
RUN curl -sSL https://get.docker.com/ | sh
This installed Docker on the container, and when I ran docker images from within the container, I could see all of the images from my host machine. I am now able to use all of the docker commands from within the container.
Related
Im trying to use the docker client from inside WSL, connecting to the docker engine on Windows. Ive exposed the docker engine on Windows on port 2375, and after setting the DOCKER_HOST environment variable in WSL, I can verify this works by running docker ps.
The problem comes when i attempt to mount directories into docker containers from WSL. For example:
I create a directory and file inside my home folder on WSL (mkdir ~/dockertest && touch ~/dockertest/example.txt)
ls ~/dockertest shows my file has been created
I now start a docker container, mounting my docker test folder (docker run -it --rm -v ~/dockertest:/data alpine ls /data)
I would expect to see 'example.txt' in the docker container, but this does not seem to be happening.
Any ideas what I might be missing?
There are great instructions for Docker setup in WSL at https://nickjanetakis.com/blog/setting-up-docker-for-windows-and-wsl-to-work-flawlessly#ensure-volume-mounts-work - solved most of my problems. The biggest trick for me was with bind-mounted directories; you have to use a path in WSL that the Docker daemon will be able to translate, for example /c/Users/rfay/myproject.
I don't think you need to change the mound point as the link suggests. Rather if you use pwd and sed in combination you should get the effect you need.
docker run -it -v $(pwd | sed 's/^\/mnt//'):/var/folder -w "/var/folder" alpine
pwd returns the working folder in the format '/mnt/c/code/folder'. Pipe this to sed and replace '/mnt' with empty string will leave you with a path such as '/c/code/folder' which is correct for docker for windows.
Anyone stumbling here over this issue follow this: Docker Desktop WSL2 Backend and make sure you are running the version 2 of the WSL in PowerShell:
> wsl -l -v
NAME STATE VERSION
* docker-desktop Running 2
Ubuntu Running 2
docker-desktop-data Running 2
If your Ubuntu VERSION doesn't say 2, you need to update it according to the guide above.
After that, you will be able to mount your Linux directory to the Docker Container directly like:
docker run -v ~/my-project:/sources <my-image>
Specific to WSL 1
Ran into the same issue. One thing to remember is that docker run command does not execute a container then and there on a command shell. It sends the run arguments to a docker daemon that does not interpret WSL path correctly. Hence, you need to pass Windows formatted path in quotes and with back slashes escaped
Your Windows path is
\\wsl$\Ubuntu\home\username\dockertest
Docker command after escaping will probably be like
docker run -it --rm -v "\\\\wsl\$\\Ubuntu\\home\\username\\dockertest":/data alpine ls /data
try
docker -v /:{path} exe
Hope to help you.
I am using jenkinssci/docker to setup some build automation on a server for a laravel project.
Using the command docker run -p 8080:8080 -p 50000:50000 -v jenkins_home:/var/jenkins_home jenkins/jenkins:lts, everything boots up fine, i create the admin login, create the project and link all of that together.
Yesterday i downloaded libraries to the container that this command gave me in docker using docker exec -u 0 -it <container_name_or_id> /bin/bash to get into the container as root to install things like php, composer, noodejs/npm. After this was done, i built the project and got a successful build.
Today I start the docker container using the same above command, build the project and build fails. The container no longer has any of the downloaded libraries (php, composer, node).
It is my understanding that including jenkins_home:/var/jenkins_home in the command to start the docker container, data would persist. This is wrong?
So my question is, how can i make it so that i can keep these libraries in the docker container that it builds?
I just started learning about these tools yesterday, so i'm not entirely sure I am even doing it the best. All i need is to be able to log into the server for Jenkins and build the project/ship the code to our staging/live servers.
side note: I am not currently using a Dockerfile. as mentioned here I am able to download tools in the container as root.
Your understanding is correct: you should use a persistent volume, otherwise you will lose your data every time the container is recreated.
I understand that you are running the container in a single machine with docker. You need to put a full path or relative path on the local folder of the volume definition to be sure that data persists, try with:
docker run -p 8080:8080 -p 50000:50000 -v ./jenkins_home:/var/jenkins_home jenkins/jenkins:lts
Look at the ./ on the local folder
Here my docker-compose.yml I'm using for a long time
version: '2'
services:
jenkins:
image: jenkins/jenkins:lts
volumes:
- ./jenkins:/var/jenkins_home
ports:
- 80:8080
- 50000:50000
Is basically the same but in yaml format
I'm brand new to both TeamCity and Docker. I'm struggling to get a Docker container with TeamCity running and usable on my local machine. I've tried several things, to no avail:
I installed Docker for Mac per instructions here. I then tried to run the following command, documented here, for setting up teamcity in docker:
docker run -it --name teamcity-server-instance \
-v c:\docker\data:/data/teamcity_server/datadir \
-v c:\docker\logs:/opt/teamcity/logs \
-p 8111:8111 \
jetbrains/teamcity-server
That returned the following error: docker: Error response from daemon: Invalid bind mount spec "c:dockerdata:/data/teamcity_server/datadir": invalid mode: /data/teamcity_server/datadir.
Taking a different tack, I tried to follow the instructions here - I tried running the following command:
docker run -it --name teamcity -p 8111:8111 sjoerdmulder/teamcity
The terminal indicated that it was starting up a web server, but I can't browse to it at localhost, nor at localhost:8111 (error ERR_SOCKET_NOT_CONNECTED without the port, and ERR_CONNECTION_REFUSED with the port).
Since the website with the docker run command says to install Docker via Docker Toolbox, I then installed that at the location they pointed to (here). I then tried the
docker-machine ip default
command they suggested, but it didn't work, error "Host does not exist: "default"". That makes sense, since the website said the "default" vm would be created by running Docker Quickstart and I didn't do that, but they don't provide any link to Docker Quickstart, so I don't know what they are talking about.
To try to get the IP address the container was running on, I tried this command
docker inspect --format='{{.Name}} - {{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' $(docker ps -aq)
That listed the names of the running containers, each followed by a hyphen, then nothing. I also tried
docker ps -a
That listed running contaners also, but didn't give the IP. Also, the port is blank, and the status says "exited (130) 4 minutes ago", so it doesn't seem like the container stayed alive after starting.
I also tried again with port 80, hoping that would make the site show at localhost:
docker run -it --name teamcity2 -p 80:80 sjoerdmulder/teamcity
So at this point, I'm completely puzzled and blocked - I can't start the server at all following the instructions on hub.docker.com, and I can't figure out how to browse to the site that does start up with the other instructions.
I'll be very grateful for any assistance!
JetBrains now provides official docker images for TeamCity. I would recommend starting with those.
The example command in their TeamCity server image looks like this
docker run -it --name teamcity-server-instance \
-v <path to data directory>:/data/teamcity_server/datadir \
-v <path to logs directory>:/opt/teamcity/logs \
-p <port on host>:8111 \
jetbrains/teamcity-server
That looks a lot like your first attempt. However, c:\docker\data is a Windows file path. You said you're running this on a mac, so that's definitely not going to work.
Once TeamCity starts, it should be available on port 8111. That's what -p 8111:8111 part of the command does. It maps port 8111 on your machine to port 8111 in the VM Docker for Mac creates to run your containers. ERR_CONNECTION_REFUSED could be caused by several things. Two most likely possibilities are
TeamCity could take a little while to start up and maybe you didn't give it enough time. Solution is to wait.
-it would start the TeamCity container in interactive mode. If you exit out of the terminal window where you ran the command, the container will also probably terminate and will be inaccessible. Solution is to not close the window or run the container in detached mode.
There is a good overview of the differences between Docker for Mac and Docker Toolbox here: Docker for Mac vs. Docker Toolbox. You don't need both, and for most cases you'll want to use Docker for Mac for testing stuff out locally.
In Docker i have installed Jenkins successfully. When i create a new job and i would like to execute a sh file from my workspace, what is the best way to add a file to my workspace with Docker? I started my container with this: docker run --name myjenkins -p 8080:8080 -p 50000:50000 -v /var/jenkins_home jenkins
You could copy a file from your file system to the container with a simple command from your terminal.
docker cp [OPTIONS] LOCALPATH|- CONTAINER:PATH
https://docs.docker.com/engine/reference/commandline/cp/
example:
docker cp /yourpaht/yourfile <containerId>:/var/jenkins_home
It depends a bit on how the planned lifecycle of your Jenkins container is. If it is just used temporarily and does no harm if the data is gone, docker cp as NickGnd suggested will do the trick.
But since the working data of Jenkins like jobconfigs, system configs and workspaces will only live inside the container, all of it will be gone once the container is removed, so if you plan to have a longer running Jenkins environment, you might want to persist the data outside of the container so it will survive recreating the container, launching new container versions and so on. This can be done with the option --volume /path/on/host:/path/in/container or its short form -v on docker run.
There is also the option of --volumes-from which you can use to mount to keep the data in one "data container" and mount it into your Jenkins container.
For further information on this, please have a look at The docker volumes documentation
I'm trying to run Docker inside a Jenkins container that is also running in Docker (i.e. Docker in Docker). What I want to know is how to properly start the Docker service when booting Jenkins. The only solution I've found today is to build my own Jenkins image based on the official Jenkins image but change the jenkins script loaded by the entry point to also start up Docker:
# I've added this line just before Jenkins is started from the script:
sudo service docker start
# I've also removed "exec" from the original file which used "exec java $JAVA_TOPS ..." but that didn't work
java $JAVA_OPTS -jar /usr/share/jenkins/jenkins.war $JENKINS_OPTS "$#"
This works when I run (using docker run) a new container but the problem is that if I do (docker start) on stopped container the Docker service is not started.
I strongly suspect that this is not the right way to start my Docker service. My plan is to perhaps use supervisord to start Jenkins and Docker separately (I suppose container linking is out of the question since Docker should be executed as a service on the same container that Jenkins is running on?). My concern with this approach is that I'm going to lose the EntryPoint specified in the Jenkins Dockerfile which allows me to pass arguments to the Jenkins container when starting the container, for example:
docker run -p 8080:8080 -v /your/home:/var/jenkins_home jenkins -- <jenkins_arguments>
Does anyone have any recommendations on a good way to solve this preferably by not forking the official Jenkins image?
I'm pretty you cannot do that.
Docker in Docker doesn't mean you have to run docker inside docker with 3 level : host > First level container > Second Level Container
In fact, you just need to share docker with host, and this is your host who will run others containers.
To do that, you have to mount volume with -v parameter
-v /var/run/docker.sock:/var/run/docker.sock
with this command, when you will docker run inside you jenkins container, the docker client will communicate with docker deamon from your host in order to run new container.
To do that, you should run your jenkins container with privileged
--privileged
To resume, here is the full command line
docker run -d -v /var/run/docker.sock:/var/run/docker.sock --privileged myimage
And you you don't need to create a new jenkins image for that.
Hoping to have helped you
http://container-solutions.com/running-docker-in-jenkins-in-docker/