I am running docker "rootless" according to this guide: https://docs.docker.com/engine/security/rootless/
The user which actually runs docker is svc_test.
When I try and start a docker container which has diretory mounts which don't exists - the docker daemon (a.k.a. svc_test user) attempts to mkdir these directories, but fails with
docker: Error response from daemon: error while creating mount source path '/dir_path/dir_name': mkdir /dir_path/dir_name: permission denied.
When I (svc_test) them attempt to do mkdir /dir_path/dir_name I succeed without any issues.
What is going on here and why does this happen?
Clearly I am missing something, but I can't trace what is that exactly.
Update 1:
This is the specific docker cmd I use to run the container:
docker run -d --restart unless-stopped \
--name questdb \
-e QDB_METRICS_ENABLED=TRUE \
--network="host" \
-v /my_mounted_volume/questdb:/questdb \
-v /my_mounted_volume/questdb/public:/questdb/public \
-v /my_mounted_volume/questdb/conf:/questdb/conf \
-v /my_mounted_volume/questdb/db:/questdb/db \
-v /my_mounted_volume/questdb/log:/questdb/log \
questdb/questdb:6.5.2 /usr/bin/env QDB_PACKAGE=docker /app/bin/java \
-m io.questdb/io.questdb.ServerMain \
-d /questdb \
-f
For clarity: my final goal is to be able to run the docker container in question from the same user form which I run my docker daemon (the svc_test user). Hence how I stumbled on this problem.
I have a Jenkins running as a docker container, now I want to build a Docker image using pipeline, but Jenkins container always tells Docker not found.
[simple-tdd-pipeline] Running shell script
+ docker build -t simple-tdd .
/var/jenkins_home/workspace/simple-tdd-pipeline#tmp/durable-
ebc35179/script.sh: 2: /var/jenkins_home/workspace/simple-tdd-
pipeline#tmp/durable-ebc35179/script.sh: docker: not found
Here is how I run my Jenkins image:
docker run --name myjenkins -p 8080:8080 -p 50000:50000 -v
/var/jenkins_home -v /var/run/docker.sock:/var/run/docker.sock
jenkins
And the DockerFile of Jenkins image is:
https://github.com/jenkinsci/docker/blob/9f29488b77c2005bbbc5c936d47e697689f8ef6e/Dockerfile
You're missing the docker client. Install it as this in Dockerfile:
RUN curl -fsSLO https://get.docker.com/builds/Linux/x86_64/docker-17.04.0-ce.tgz \
&& tar xzvf docker-17.04.0-ce.tgz \
&& mv docker/docker /usr/local/bin \
&& rm -r docker docker-17.04.0-ce.tgz
Source
In your Jenkins interface go to "Manage Jenkins/Global Tool Configuration"
Then scroll down to Docker Installations and click "Add Docker". Give it a name like "myDocker"
Make sure to check the box which says "Install automatically". Click "Add Installer" and select "Download from docker.com". Leave "latest" in the Docker version. Make sure you click Save.
In your Jenkinsfile add the following stage before you run any docker commands:
stage('Initialize'){
def dockerHome = tool 'myDocker'
env.PATH = "${dockerHome}/bin:${env.PATH}"
}
Edit: May 2018
As pointed by Guillaume Husta, this jpetazzo's blog article discourages this technique:
Former versions of this post advised to bind-mount the docker binary from the host to the container. This is not reliable anymore, because the Docker Engine is no longer distributed as (almost) static libraries.
Docker client should be installed inside a container as described here. Also, jenkins user should be in docker group, so execute following:
$ docker exec -it -u root my-jenkins /bin/bash
# usermod -aG docker jenkins
and finally restart my-jenkins container.
Original answer:
You could use host's docker engine like in this #Adrian Mouat blog article.
docker run -d \
--name my-jenkins \
-v /var/jenkins_home:~/.jenkins \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 8080:8080 jenkins
This avoids having multiple docker engine version on host and jenkins container.
The problem is in your Jenkins, it isn't capable to use the docker engine, even if you do install the docker from the plugin manager. From what I got researching there are some alternatives to workaround this issue:
1: Build a image using some docker image with pre-installed docker in it like provided by getintodevops/jenkins-withdocker:lts
2: Build the images from jenkins/jenkins mounting the volumes to your host then install the docker all by yourself by creating another container with same volumes and executing the bash cmd to install the docker or using Robert suggestion
docker run -p 8080:8080 -p 50000:50000 -v $HOME/.jenkins/:/var/jenkins_home
-v /var/run/docker.sock:/var/run/docker.sock jenkins/jenkins:latest
or 3: The most simple, just add the installed docker path from your host machine to be used by your jenkins container with: -v $(which docker):/usr/bin/docker
Your docker command should look like this:
docker run \
--name jenkins --rm \
-u root -p 8080:8080 -p 50000:50000 \
-v $(which docker):/usr/bin/docker\
-v $HOME/.jenkins/:/var/jenkins_home
-v /var/run/docker.sock:/var/run/docker.sock \
jenkins/jenkins:latest
[Source]https://forums.docker.com/t/docker-not-found-in-jenkins-pipeline/31683
Extra option: Makes no sense if you just want to make use of a single Jenkis server but it's always possible to install a OS like Ubuntu using an image and install the jenkins .war file from there
docker run -d \
--group-add docker \
-v $(pwd)/jenkins_home:/var/jenkins_home \
-v /var/run/docker.sock:/var/run/docker.sock \
-v $(which docker):/usr/bin/docker \
-p 8080:8080 -p 50000:50000 \
jenkins/jenkins:lts
Just add option --group-add docker when docker run.
Add docker path i.e -v $(which docker):/usr/bin/docker to container in volumes like
docker run -d \
--name my-jenkins \
-v $(which docker):/usr/bin/docker \
-v /var/jenkins_home:~/.jenkins \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 8080:8080 jenkins
This section helped me install docker inside the jenkins container: https://www.jenkins.io/doc/book/installing/docker/#downloading-and-running-jenkins-in-docker
Also, I had to replace FROM jenkins/jenkins:2.303.1-lts-jdk11 in the Dockerfile in step 4(a) with jenkins/jenkins.
I am setting up quay in a vm with centos distro. This is the guide I am following: quay deploy guide
once I install Podman I am trying to run first container with below command:
I set up this env variable:
export QUAY=QUAY
and made a dir of same name in home:
mkdir QUAY
once I install Podman I am trying to run first container with below command:
$ sudo podman run -d --rm --name postgresql-quay \
-e POSTGRESQL_USER=quayuser \
-e POSTGRESQL_PASSWORD=quaypass \
-e POSTGRESQL_DATABASE=quay \
-e POSTGRESQL_ADMIN_PASSWORD=adminpass \
-p 5432:5432 \
-v $QUAY/postgres-quay:/var/lib/pgsql/data:Z \
registry.redhat.io/rhel8/postgresql-10:1
and I am getting following error:
sudo podman run -d --rm --name postgresql-quay -e POSTGRESQL_USER=quayuser -e POSTGRESQL_PASSWORD=quaypass -e POSTGRESQL_DATABASE=quay -e POSTGRESQL_ADMIN_PASSWORD=adminpass -p 5432:5432 -v QUAY/postgres-quay:/var/lib/pgsql/data:Z registry.redhat.io/rhel8/postgresql-10:1
Error: error creating named volume "QUAY/postgres-quay": error running volume create option: names must match [a-zA-Z0-9][a-zA-Z0-9_.-]*: invalid argument
The bind mount needs to be specified as an absolute path or a relative path that starts with ./ or ../.
In other words instead of
-v QUAY/postgres-quay:/var/lib/pgsql/data:Z
use
-v ./QUAY/postgres-quay:/var/lib/pgsql/data:Z
(I replaced $QUAY with its value QUAY)
I ran the docker run command to run the jenkins image.
But this is the command that was executed in the past, but now an error occurs.
command
docker run -u root —rm -p 8080:8080 —name jenkins \ -v Users/human/jenkins:/var/jenkins_home \
-v /var/run/docker.sock:/var/run/docker.sock \ jenkins
error
docker: invalid reference format.
See 'docker run --help'.
My environment is Mac OS X. What's wrong with me?
It could be the path "Users/human/jenkins" is invalid. If you are using $PWD following the example on Jenkins site, be very careful about possible space character in the "current" path. That would mess up the command and throw out that error message you posted.
It seems like you have a parsing issue for the command above, I have changed this — to --.. this should work:
docker run -u root --rm -p 8080:8080 --name jenkins -v Users/human/jenkins:/var/jenkins_home -v /var/run/docker.sock:/var/run/docker.sock jenkins
Or like this:
docker run -u root --rm -p 8080:8080 --name jenkins \
-v Users/human/jenkins:/var/jenkins_home \
-v /var/run/docker.sock:/var/run/docker.sock \
jenkins
Really new to docker and docker infrastructure. So from the official Zalenium github page I followed all the steps successfully until I get to the point:
docker run --rm -ti --name zalenium -p 4444:4444 \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /tmp/videos:/home/seluser/videos \
--privileged dosel/zalenium start
and after executing it I'm getting:
> docker: invalid reference format.
> See 'docker run --help'
While doing research here on SOF, I've landed here and here, but nothing seems to work in my situation. I've tried to replace --rm -ti to the end of the expression: didn't work
docker run --name zalenium -p 4444:4444 \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /tmp/videos:/home/seluser/videos \
--privileged --rm -ti dosel/zalenium start
I'm sure there is a specific format that I have to follow, but can't find a clue to it (missing comma, wrong backslash, unnecessary whitespace). Appreciate for your help.
I just noticed that the command example we have does not work for Windows, I just changed the docs to show how to run a multiline command in Windows.
Here is the way:
docker run --rm -ti --name zalenium -p 4444:4444 ^
-v /var/run/docker.sock:/var/run/docker.sock ^
-v /c/Users/your_user_name/temp/videos:/home/seluser/videos ^
--privileged dosel/zalenium start