settings.bash

#!/bin/bash

# set a project name, no empty spaces or special characters allowed
export PROJECT_NAME=docker_image_development

# path to your docker registry, leave blank if you don't have one
# e.g. my.registry.com, ghcr.io/dfki-ric, docker.pkg.github.com, harbor.mydomain.com
export DOCKER_REGISTRY=
# When your registry has subfolders/groups you can add it here 
export DOCKER_REGISTRY_GROUP=developmentimage

# in case you are not using a single registry, you can push images in different ones
# e.g. store base images on hub.docker.com and others in a local registry
export BASE_REGISTRY=
export DEVEL_REGISTRY=$DOCKER_REGISTRY
export RELEASE_REGISTRY=$DOCKER_REGISTRY

# when your images should be stored in different groups of your registry, you can chage it here
# (or completely customize by setting WORKSPACE_DEVEL_IMAGE, WORKSPACE_RELEASE_IMAGE, WORKSPACE_CD_IMAGE below directly)
export DEVEL_REGISTRY_GROUP=$DOCKER_REGISTRY_GROUP
export RELEASE_REGISTRY_GROUP=$DOCKER_REGISTRY_GROUP
export CD_REGISTRY_GROUP=$DOCKER_REGISTRY_GROUP

# should exec and build scripts auto-pull updated images from the registry?
export DOCKER_REGISTRY_AUTOPULL=false

### The default release mode to use if no mode paramater is given to ./exec.bash or ./stop.bash
### The checked in version should reflect the image status and be the highest availale image (base - devel - release)
export DEFAULT_EXECMODE="base" # Use this only for setting up the initial devel image (modify setup_workspace.bash)
# export DEFAULT_EXECMODE="devel" # This is used while deveoping code and preparing a relase
# export DEFAULT_EXECMODE="release" # use the release as default
# export DEFAULT_EXECMODE="CD" # use the continuous deployment image as default

### The base image used when building a workspace image (one of the ones build in base_images)
# export WORKSPACE_BASE_IMAGE=developmentimage/rock_master_20.04:base # image with rock core dependencies installed
# export WORKSPACE_BASE_IMAGE=developmentimage/rock_master_22.04:base # image with rock core dependencies installed
# export WORKSPACE_BASE_IMAGE=developmentimage/ros_noetic_20.04:base # image with basic ros noetic installed
# export WORKSPACE_BASE_IMAGE=developmentimage/ros2_foxy_20.04:base # image with ros2 foxy desktop installed
# export WORKSPACE_BASE_IMAGE=developmentimage/ros2_humble_22.04:base # image with ros2 humble desktop installed
# export WORKSPACE_BASE_IMAGE=developmentimage/plain_20.04:base # plain image with build_essentials installed
export WORKSPACE_BASE_IMAGE=developmentimage/plain_22.04:base # plain image with build_essentials installed
# export WORKSPACE_BASE_IMAGE=developmentimage/plain_24.04:base # plain image with build_essentials installed

# The Name of the Workspace image to use
# you should add a workspace name folder and a image name
# e.g MY_PROJECT/docker_image_development:devel
# under normal circumstances you should not need to change the values here
export WORKSPACE_DEVEL_IMAGE=${DEVEL_REGISTRY_GROUP}/${PROJECT_NAME}:devel
export WORKSPACE_RELEASE_IMAGE=${RELEASE_REGISTRY_GROUP}/${PROJECT_NAME}:release
export WORKSPACE_CD_IMAGE=${CD_REGISTRY_GROUP}/${PROJECT_NAME}:CD

# In case your docker container needs special run paramaters
# like open ports, additional mounts etc.
# When you change this, you need to recreate the container
# best way to do this, run the delete_contianer.bash script
# often used params:
# --dns-search=mydomain
# --net=host
# --privileged
export ADDITIONAL_DOCKER_RUN_ARGS=""

# Make the exec script to talk more for debugging/docker setup purposes.
# This may also be stated in the command line: $> VERBOSE=true ./exec.bash 
# export VERBOSE=true

# Make the output as quiet as possible (does not apply to programs started in the container)
# export SILENT=false

# make the local ssh-agent available in the container
# with this you can checkout code into your devel contaienr usind ssh
# export MOUNT_SSH_AGENT=true

# mount ccache volume, if enabled, a volume name based on the base image name is generated
# and mounted to /ccache, this way multiple workspaces in docker_image_development
# can share a single ccache, CCACHE_DIR is automatically set in the env, just install
# and enable ccache for your builds
# export MOUNT_CCACHE_VOLUME=true

# Icecc will only work in a single container at a time.
# For icecc to work, you need to make int available from the host, so either use
# "--net=host" or "-p 10245:10245 -p 8765:8765/udp" in the ADDITIONAL_DOCKER_RUN_ARGS.
# if none of them is added there, "-p 10245:10245 -p 8765:8765/udp" will be added automatically with a warning
# WARNING: using "-p 10245:10245 -p 8765:8765/udp" will block the port on the host, only allowing one container to exist at a time using the ports.
# stop the container using the port before launching the next one, or use --net=host for the containers and stop the iceccd service.
# You'll have to enable the use of icecc for your workspace manually, this only set up the availability of icecc in the container
# export USE_ICECC=true

# conenct to xserver via [mount, xpra, none, auto]
# auto will detect if the container is started over an ssh session and if yes, xpra is used, mount otherwise
export DOCKER_XSERVER_TYPE=auto
#xpra_port may be set if --net=host is used, otherwise, please use -p in the ADDITIONAL_DOCKER_RUN_ARGS to assign a port for the
#xpra server, DOCKER_XSERVER_TYPE needs to be "xpra"
export XPRA_PORT="10000"

# always update DISPLAY variable for new ./exec.bash commands
# when ./exec.bash is called through ssh -X each ssh conenction will have its own DISPLAY)
# this also needs --net=host (auto-added) to reach the xserver via localhost
export USE_XSERVER_VIA_SSH=false


# If you need to start docker contaiers from your workspace (e.g. for launch tools) set this option to true
# It will add "--privileged -v /var/run/docker.sock:/var/run/docker.sock" to the DOCKER_RUN_ARGS
# The hosts docker deamon can then be used from within the container. Setting this to true will also add the dockeruser
# to the docker group inside the container and set the docker groups id in the container to have the same gid as the host docker group.
# The docker.io (apt) package has to be installed manually (e.g. through the 02_devel_image/Dockerfile)
# Also, add "RUN adduser dockeruser docker" that file, if you want to use it with the default user
export NEEDS_DOCKER_IN_CONTAINER=false


# To allow profiling with perf tools on the host, mount the container roor dir to a local folder (./container_root)
# Then run profiling tools on the host, but provide root folder location to search for libs:
# sudo hotspot --sysroot ./container_root
# sudo perf record -o perf.data --call-graph dwarf --aio -z --pid $(pidof my_program) && perf report --symfs ./container_root
# needs bindfs installed on the host
export MOUNT_CONTAINER_ROOT=false