Skip to content
This repository has been archived by the owner on Oct 10, 2020. It is now read-only.
/ atomic Public archive

Atomic Run Tool for installing/running/managing container images.

License

Unknown, Unknown licenses found

Licenses found

Unknown
LICENSE
Unknown
COPYING
Notifications You must be signed in to change notification settings

projectatomic/atomic

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

⚠️ The Atomic tool has been deprecated. ⚠️

The Podman project has replaced it with most of its functionality implemented. Please open issues and questions with the Podman repository.

Atomic: /usr/bin/atomic

This project defines the entrypoint for Project Atomic hosts. On an Atomic Host, there are at least two distinct software delivery vehicles; Docker (often used in combination with the traditional RPM/yum/dnf), and rpm-ostree to provide atomic upgrades of the host system.

The goal of Atomic is to provide a high level, coherent entrypoint to the system, and fill in gaps in Linux container implementations.

For Docker, atomic can make it easier to interact with special kinds of containers, such as super-privileged debugging tools and the like.

The atomic host subcommand wraps rpm-ostree, currently just providing a friendlier name, but in the future Atomic may provide more unified management.

atomic run

Atomic allows an image provider to specify how a container image expects to be run.

Specifically this includes the privilege level required.

For example if you built an 'ntpd' container application, that required the SYS_TIME capability, you could add meta data to your container image using the command:

LABEL RUN /usr/bin/docker run -d --cap-add=SYS_TIME ntpd

Now if you executed atomic run ntpd, it would read the LABEL RUN json metadata from the container image and execute this command.

atomic install

Most of the time when you ship an application, you need to run an install script. This script would configure the system to run the application, for example it might configure a systemd unit file or configure kubernetes to run the application. This tool will allow application developers to embed the install and uninstall scripts within the application. The application developers can then define the LABEL INSTALL and LABEL UNINSTALL methods, in the image meta data. Here is a simple httpd installation description.

cat Dockerfile

# Example Dockerfile for httpd application
#
FROM		fedora
MAINTAINER	Dan Walsh
ENV container docker
RUN yum -y update; yum -y install httpd; yum clean all

LABEL Vendor="Red Hat" License=GPLv2
LABEL Version=1.0
LABEL INSTALL="docker run --rm --privileged -v /:/host -e HOST=/host -e LOGDIR=/var/log/\${NAME} -e CONFDIR=/etc/\${NAME} -e DATADIR=/var/lib/\${NAME} -e IMAGE=\${IMAGE} -e NAME=\${NAME} \${IMAGE} /bin/install.sh"
LABEL UNINSTALL="docker run --rm --privileged -v /:/host -e HOST=/host -e IMAGE=${IMAGE} -e NAME=${NAME} ${IMAGE} /bin/uninstall.sh"
ADD root /

EXPOSE 80

CMD [ "/usr/sbin/httpd", "-D", "FOREGROUND" ]

atomic install will read the LABEL INSTALL line and substitute ${NAME} with the name specified with the name option, or use the image name, it will also replace${IMAGE} with the image name.

To be used by the application. The install script could populate these directories if necessary.

In my example the INSTALL method will execute the install.sh which we add to the image. The root sub directory contains the following scripts:

The atomic install will set the following environment variables for use in the command:

SUDO_UID The SUDO_UID environment variable. This is useful with the docker -u option for user space tools. If the environment variable is not available, the value of /proc/self/loginuid is used.

SUDO_GID The SUDO_GID environment variable. This is useful with the docker -u option for user space tools. If the environment variable is not available, the default GID of the value for SUDO_UID is used. If this value is not available, the value of /proc/self/loginuid is used.

cat root/usr/bin/install.sh

#!/bin/sh
# Make Data Dirs
mkdir -p ${HOST}/${CONFDIR} ${HOST}/${LOGDIR}/httpd ${HOST}/${DATADIR}

# Copy Config
cp -pR /etc/httpd ${HOST}/${CONFDIR}

# Create Container
chroot ${HOST} /usr/bin/docker create -v /var/log/${NAME}/httpd:/var/log/httpd:Z -v /var/lib/${NAME}:/var/lib/httpd:Z --name ${NAME} ${IMAGE}

# Install systemd unit file for running container
sed -e "s/TEMPLATE/${NAME}/g" etc/systemd/system/httpd_template.service > ${HOST}/etc/systemd/system/httpd_${NAME}.service

# Enabled systemd unit file
chroot ${HOST} /usr/bin/systemctl enable /etc/systemd/system/httpd_${NAME}.service

atomic uninstall

The atomic unistall does the same variable substitution as described for install, and can be used to remove any host system configuration.

Here is the example script we used.

cat root/usr/bin/uninstall.sh

#!/bin/sh
chroot ${HOST} /usr/bin/systemctl disable /etc/systemd/system/httpd_${NAME}.service
rm -f ${HOST}/etc/systemd/system/httpd_${NAME}.service

Finally here is the systemd unit file template I used:

cat root/etc/systemd/system/httpd_template.service

# cat ./root/etc/systemd/system/httpd_template.service
[Unit]
Description=The Apache HTTP Server for TEMPLATE
After=docker.service
BindTo=docker.service

[Service]
ExecStart=/usr/bin/docker start TEMPLATE
ExecStop=/usr/bin/docker stop TEMPLATE
ExecReload=/usr/bin/docker exec -t TEMPLATE /usr/sbin/httpd $OPTIONS -k graceful

[Install]
WantedBy=multi-user.target

For an explaination of the Atomic scan JSON output, see the JSON specification document.