Update and fix gzdev jobs in Jenkins (#1128)

* Modernize DSL distributions
* Update and fix the docker-in-docker support until the point of being able to run rocker
* Remove old docker install code
---------

Signed-off-by: Jose Luis Rivero <jrivero@osrfoundation.org>

jenkins-scripts/docker/lib/_install_nvidia_docker.sh

@@ -1,22 +1,7 @@
-INSTALL_NVIDIA_DOCKER1="""
-echo '# BEGIN SECTION: install docker (in docker)'
-curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
-sudo add-apt-repository \"deb [arch=amd64] https://download.docker.com/linux/ubuntu \$(lsb_release -cs) stable\"
-sudo apt-get update
-sudo apt-get install -y docker-ce
-echo '# END SECTION'
-
-echo '# BEGIN SECTION: install nvidia-docker1 (in docker)'
-sudo apt-get install -y wget nvidia-340 nvidia-modprobe
-wget -P /tmp https://github.com/NVIDIA/nvidia-docker/releases/download/v1.0.1/nvidia-docker_1.0.1-1_amd64.deb
-sudo dpkg -i /tmp/nvidia-docker*.deb && rm /tmp/nvidia-docker*.deb
-echo '# END SECTION'
-"""
-
 INSTALL_NVIDIA_DOCKER2="""
 echo '# BEGIN SECTION: install docker (in docker)'
 curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
-sudo add-apt-repository \"deb [arch=amd64] https://download.docker.com/linux/ubuntu \$(lsb_release -cs) stable\"
+sudo add-apt-repository -y \"deb [arch=amd64] https://download.docker.com/linux/ubuntu \$(lsb_release -cs) stable\"
 sudo apt-get update
 # sudo apt-get install -y docker-ce
 echo '# END SECTION'
@@ -29,11 +14,8 @@ curl -s -L https://nvidia.github.io/nvidia-docker/\$distribution/nvidia-docker.l
   sudo tee /etc/apt/sources.list.d/nvidia-docker.list
 sudo apt-get update
 sudo apt-get install -y nvidia-docker2
-# systemctl daemon-reload
-# systemctl restart docker
+sudo docker info
 echo '# END SECTION'
 """
 
-DOCKER2_CMD="sudo docker run --privileged -e DISPLAY=unix:0 -v /sys:/sys:ro -v /var/run/docker.sock:/var/run/docker.sock -v /tmp/.X11-unix:/tmp/.X11-unix:rw --runtime=nvidia -e DOCKER_FIX= -v /dev/log:/dev/log:ro -v /run/log:/run/log:ro -v /sys/fs/cgroup:/sys/fs/cgroup:ro --device /dev/snd --tty --rm"
-
 DEPENDENCY_PKGS="${DEPENDENCY_PKGS} curl software-properties-common"

jenkins-scripts/docker/lib/docker_generate_dockerfile.bash

@@ -413,6 +413,13 @@ DELIM_DISPLAY
   fi
 fi
 
+if ${USE_DOCKER_IN_DOCKER}; then
+cat >> Dockerfile << DELIM_WORKAROUND_DOCKER_IN_DOCKER_HOOK
+# Avoid ERROR: invoke-rc.d: policy-rc.d denied execution of start.
+RUN sed -i "s/^exit 101$/exit 0/" /usr/sbin/policy-rc.d
+DELIM_WORKAROUND_DOCKER_IN_DOCKER_HOOK
+fi
+
 if [ `expr length "${DOCKER_POSTINSTALL_HOOK}"` -gt 1 ]; then
 cat >> Dockerfile << DELIM_WORKAROUND_POST_HOOK
 RUN ${DOCKER_POSTINSTALL_HOOK}
@@ -468,11 +475,16 @@ RUN adduser --uid \$USERID --gid \$GID --gecos "Developer" --disabled-password \
 RUN adduser \$USER sudo
 RUN echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers
 RUN chown -R \$USER:\$USER /home/\$USER
+# Needed if USE_DOCKER_IN_DOCKER is active. Harmless to be here
+RUN groupadd docker
+RUN gpasswd -a \$USER docker
+RUN newgrp docker
 
 # permit access to USER variable inside docker
 ENV USER \$USER
 USER \$USER
 # Must use sudo where necessary from this point on
+
 DELIM_DOCKER_USER
 
 if [[ -n ${SOFTWARE_DIR} ]]; then

jenkins-scripts/docker/lib/gzdev-base-linux.bash

@@ -13,49 +13,49 @@ set -ex
 export MAKE_JOBS=${MAKE_JOBS}
 export DISPLAY=${DISPLAY}
 
-${INSTALL_NVIDIA_DOCKER1}
+${INSTALL_NVIDIA_DOCKER2}
 
 echo '# BEGIN SECTION: install pip requirements'
 cd ${WORKSPACE}/gzdev
-pip3 install -r requirements.txt
+# pip3 install -r requirements.txt
 echo '# END SECTION'
 
-echo '# BEGIN SECTION: run gzdev for gazebo8 with nvidia'
-cd ${WORKSPACE}/gzdev
-./gzdev.py spawn --gzv=8 --nvidia
-echo '# END SECTION'
+echo '# BEGIN SECTION: smoke tests for ign-docker-env'
+sudo pip3 install git+https://github.com/adlarkin/ign-rocker.git
 
-echo '# BEGIN SECTION: Disply log file.'
-cat ./gz8.log
-echo '# END SECTION'
+# TODO: rocker can not play well docker-in-docker installations
+# out of the box. Needs more work.
+
+# xvfb :1 -ac -noreset -core -screen 0 1280x1024x24 &
+# export display=:1.0
+# export mesa_gl_version_override=3.3
+
+# test_timeout=300
+# test_start=\`date +%s\`
+# sudo bash -c "timeout --preserve-status \$test_timeout ./gzdev.py ign-docker-env dome --linux-distro ubuntu:bionic"
+# test_end=\`date +%s\`
+# diff=\$(expr \$test_end - \$test_start)
 
-echo '# BEGIN SECTION: check that gazebo is running'
-gazebo_detection=false
-seconds_waiting=0
-cat gz8.log | grep "Connected to gazebo master" && gazebo_detection=true
-while (! \$gazebo_detection); do
-   sleep 1
-   docker top gz8 | grep gazebo && gazebo_detection=true
-   docker top gz8 | grep gzserver && gazebo_detection=true
-   seconds_waiting=\$((seconds_waiting+1))
-   [ \$seconds_waiting -gt 30 ] && break
-done
-# clean up gazebo instances
-docker rm -f gz8 || true
-killall -9 gazebo gzserver gzclient || true
-! \${gazebo_detection} && exit 1
+# if [ \$diff -lt \$test_timeout ]; then
+#    echo 'the test took less than \$test_timeout. something bad happened'
+#    exit 1
+# fi
 echo '# END SECTION'
 DELIM
 
 export USE_DOCKER_IN_DOCKER=true
 export OSRF_REPOS_TO_USE="stable"
+export USE_ROS_REPO=true
+export ROS2=true
 export DEPENDENCY_PKGS="python3-pip \
                  bash \
                  apt-transport-https \
                  ca-certificates \
                  curl \
                  software-properties-common \
-		 psmisc" # killall
+                 python3-rocker \
+                 psmisc \
+                 xvfb"
 
 . "${SCRIPT_DIR}/lib/docker_generate_dockerfile.bash"
 . "${SCRIPT_DIR}/lib/docker_run.bash"

jenkins-scripts/dsl/gzdev.dsl

@@ -1,7 +1,7 @@
 import _configs_.*
 import javaposse.jobdsl.dsl.Job
 
-def supported_distros = [ 'bionic' ]
+def supported_distros = [ 'jammy' ]
 def supported_arches = [ 'amd64' ]
 
 
@@ -17,6 +17,8 @@ supported_distros.each { distro ->
 
     gzdev_ci_job.with
     {
+
+      label Globals.nontest_label("large-memory")
 
       scm {
         git {
@@ -54,7 +56,6 @@ supported_distros.each { distro ->
 
     gzdev_any_job.with
     {
-      // use only the most powerful nodes
       label Globals.nontest_label("large-memory")
 
       steps {