Handy Docker Scripts

Introduction:

Here we go, yet another docker post!

I'm going to keep this one short and simply use it as a repository to store some bash scripts I've written to automate some docker functions.

Scripts:

Some of these are rather hacked together, they generally work but may need some tweaking here and there!

Clean Images:

This script removes redundant images which are marked as dangling by the docker engine.

#!/bin/bash
#Author: David Chidell

echo '***Removing redundant images...'  
IMAGES=$(docker images --filter dangling=true -q)  
docker rmi $IMAGES > /dev/null 2>&1  

Removed unused images:

This script compares the output of docker ps -a and docker images and finds images which have been downloaded but not used for any existing containers, then removes the containers which are unused. Useful for cleaning up after playing around.

#!/bin/bash
#Author: David Chidell

echo '***Removing unused images...'

used_images=$(docker ps -a | tail -n +2 | awk {'print $2'} | sort | uniq)  
all_images=$(docker images | grep -v REPOSITORY | cut -d' ' -f1 | sort | uniq)  
unused_images=$(diff <(echo "$all_images") <(echo "$used_images") | grep '<' | awk {'print $2'})  
docker rmi $unused_images 2>&1  

Update containers / apps:

This one is a little more full-on. It's also rather insecure but we don't worry about things like that!

This script performs the following:

• Downloads the latest image for running containers
• Finds container definitions contained within the containers directory.
  • Deletes the old container
  • Re-creates the new container (more on container definitions later)
  • Starts the new container
• Finds app definitions (based on docker-compose) contained within the apps directory.
  • Deletes all containers part of the app
  • Re-creates the containers based on the docker-compose file
  • Starts the app
• Deletes old redundant images

#!/bin/bash
#Author: David Chidell

echo '***Updating Images...'  
#This line updates ALL images in the repo
#docker images | grep -v REPOSITORY | cut -d' ' -f1 | xargs -L1 docker pull 
#This one only updates images we're using.
docker ps | tail -n +2 | awk {'print $2'} | sort | uniq | xargs -L1 docker pull 

echo '***Updating Containers...'  
CONTAINERS=$(ls containers/ | grep .txt)  
for container in $CONTAINERS  
do  
 container_name=$(echo $container | cut -d'.' -f1)
 echo '*Stopping and removing ' $container_name
 docker stop $container_name > /dev/null 2>&1
 docker rm $container_name > /dev/null 2>&1
 echo '*Creating and starting ' $container_name
 docker start $(bash containers/$container) > /dev/null 2>&1
done

echo '***Updating Applications...'  
DIR=$pwd  
APPS=$(ls apps/ | grep .txt)  
for app in $APPS  
do  
 cd apps/$app
 docker-compose stop
 docker-compose rm -f
 docker-compose up -d
 cd $DIR
done

#Cleanup redundant images
bash $DIR/clean_images.sh  

Container and application definitions are stored as .txt files inside the containers and apps directories respectively.

These definitions are essentially bash commands which would be used to create a container. An example of a container.txt file would be as follows:

docker create \  
--name=portainer \
--restart=always \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 9000:9000 \
portainer/portainer  

This is generally insecure as this file is executed by bash. These files can be manipulated to contain any bash script.

A container / app can be disabled by changing the file extension from .txt to something else (Note: If the file still contains the .txt extension it will NOT be disabled!)