From 4c01651f3084875b67d2b2690e3cc555b3b85665 Mon Sep 17 00:00:00 2001
From: Luca Cinquini <luca.cinquini@jpl.nasa.gov>
Date: Tue, 27 Feb 2024 07:05:44 -0700
Subject: [PATCH 1/3] feat(): Add build of Docker image to execute CWL
 workflows

---
 .github/workflows/build_docker_images.yml   | 24 +++++++++++
 airflow/docker/cwl/Dockerfile               | 25 ++++++++++++
 airflow/docker/cwl/docker_cwl_entrypoint.sh | 44 +++++++++++++++++++++
 3 files changed, 93 insertions(+)
 create mode 100644 airflow/docker/cwl/Dockerfile
 create mode 100755 airflow/docker/cwl/docker_cwl_entrypoint.sh

diff --git a/.github/workflows/build_docker_images.yml b/.github/workflows/build_docker_images.yml
index 80090b42..e940b3a6 100644
--- a/.github/workflows/build_docker_images.yml
+++ b/.github/workflows/build_docker_images.yml
@@ -12,6 +12,7 @@ env:
   REGISTRY: ghcr.io
   TAG: ${{ github.event.inputs.tag }}
   SPS_AIRFLOW: ${{ github.repository }}/sps-airflow
+  SPS_DOCKER_CWL: ${{ github.repository }}/sps-docker-cwl
 
 jobs:
   build-sps-airflow:
@@ -37,3 +38,26 @@ jobs:
           push: true
           tags: ${{ env.REGISTRY }}/${{ env.SPS_AIRFLOW }}:${{ env.TAG }}
           labels: ${{ steps.metascheduler.outputs.labels }}
+  build-sps-docker-cwl:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - name: Log in to the Container registry
+        uses: docker/login-action@f054a8b539a109f9f41c372932f1ae047eff08c9
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+      - name: Extract metadata (tags, labels) for SPS Docker CWL image
+        id: metascheduler
+        uses: docker/metadata-action@98669ae865ea3cffbcbaa878cf57c20bbf1c6c38
+        with:
+          images: ${{ env.REGISTRY }}/${{ env.SPS_DOCKER_CWL }}
+      - name: Build and push SPS Docker CWL image
+        uses: docker/build-push-action@ad44023a93711e3deb337508980b4b5e9bcdc5dc
+        with:
+          context: ./airflow/docker/cwl
+          file: airflow/docker/cwl/Dockerfile
+          push: true
+          tags: ${{ env.REGISTRY }}/${{ env.SPS_DOCKER_CWL }}:${{ env.TAG }}
+          labels: ${{ steps.metascheduler.outputs.labels }}
diff --git a/airflow/docker/cwl/Dockerfile b/airflow/docker/cwl/Dockerfile
new file mode 100644
index 00000000..781a95b1
--- /dev/null
+++ b/airflow/docker/cwl/Dockerfile
@@ -0,0 +1,25 @@
+# docker:dind Dockerfile: https://github.com/docker-library/docker/blob/master/Dockerfile-dind.template
+# FROM docker:dind
+FROM docker:25.0.3-dind
+
+# install Python
+RUN apk add --update --no-cache python3 && ln -sf python3 /usr/bin/python
+RUN apk add gcc musl-dev linux-headers python3-dev
+RUN apk add --no-cache python3 py3-pip
+RUN apk add vim
+
+# install CWL libraries
+RUN mkdir /usr/share/cwl \
+    && cd /usr/share/cwl \
+    && python -m venv venv \
+    && source venv/bin/activate \
+    && pip install cwltool cwl-runner docker boto3 awscli
+
+# install nodejs to parse Javascript in CWL files
+RUN apk add --no-cache nodejs npm
+
+# script to execute a generic CWL workflow with arguments
+COPY docker_cwl_entrypoint.sh /usr/share/cwl/docker_cwl_entrypoint.sh
+
+WORKDIR /usr/share/cwl
+ENTRYPOINT ["/usr/share/cwl/docker_cwl_entrypoint.sh"]
diff --git a/airflow/docker/cwl/docker_cwl_entrypoint.sh b/airflow/docker/cwl/docker_cwl_entrypoint.sh
new file mode 100755
index 00000000..6ea15517
--- /dev/null
+++ b/airflow/docker/cwl/docker_cwl_entrypoint.sh
@@ -0,0 +1,44 @@
+#!/bin/sh
+# Script to execute a CWL workflow that includes Docker containers
+# The Docker engine is started before the CWL execution, and stopped afterwards.
+# $1: the CWL workflow URL (example: https://raw.githubusercontent.com/unity-sds/unity-sps-prototype/cwl-docker/cwl/cwl_workflows/echo_from_docker.cwl)
+# $2: the CWL job parameters as a JSON formatted string (example: { name: John Doe })
+# $3: optional output directory, defaults to the current directory
+# Note: $output_dir must be accessible by the Docker container that executes this script
+
+set -ex
+cwl_workflow=$1
+job_args=$2
+output_dir=${3:-.}
+echo "Executing CWL workflow: $cwl_workflow with json arguments: $job_args and output directory: $output_dir"
+echo "$job_args" > /tmp/job_args.json
+cat /tmp/job_args.json
+
+# create output directory if it doesn't exist
+mkdir -p "$output_dir"
+
+# Start Docker engine
+dockerd > dockerd-logfile 2>&1
+
+# Wait until Docker engine is running
+# Loop until 'docker version' exits with 0.
+until docker version > /dev/null 2>&1
+do
+  sleep 1
+done
+echo "done sleeping"
+
+
+# Execute CWL workflow
+. /usr/share/cwl/venv/bin/activate
+aws s3 ls
+aws ssm get-parameter --name /sps/processing/workflows/edl_username --region us-west-2
+
+# wait for 60 minutes
+sleep 3600
+
+cwl-runner --outdir "$output_dir" --no-match-user --no-read-only "$cwl_workflow" /tmp/job_args.json
+deactivate
+
+# Stop Docker engine
+pkill -f dockerd

From b1ee5e22e7fdaac196447b7012f0683a15475c92 Mon Sep 17 00:00:00 2001
From: Luca Cinquini <luca.cinquini@jpl.nasa.gov>
Date: Tue, 27 Feb 2024 07:12:58 -0700
Subject: [PATCH 2/3] fix: cleaning up the entrypoint script

---
 airflow/docker/cwl/docker_cwl_entrypoint.sh | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/airflow/docker/cwl/docker_cwl_entrypoint.sh b/airflow/docker/cwl/docker_cwl_entrypoint.sh
index 6ea15517..da2aacc4 100755
--- a/airflow/docker/cwl/docker_cwl_entrypoint.sh
+++ b/airflow/docker/cwl/docker_cwl_entrypoint.sh
@@ -26,17 +26,9 @@ until docker version > /dev/null 2>&1
 do
   sleep 1
 done
-echo "done sleeping"
-
 
 # Execute CWL workflow
 . /usr/share/cwl/venv/bin/activate
-aws s3 ls
-aws ssm get-parameter --name /sps/processing/workflows/edl_username --region us-west-2
-
-# wait for 60 minutes
-sleep 3600
-
 cwl-runner --outdir "$output_dir" --no-match-user --no-read-only "$cwl_workflow" /tmp/job_args.json
 deactivate
 

From a89e1edfddb8596b4f9991b70880df2fc1b39500 Mon Sep 17 00:00:00 2001
From: Luca Cinquini <luca.cinquini@jpl.nasa.gov>
Date: Tue, 27 Feb 2024 07:35:45 -0700
Subject: [PATCH 3/3] fix: the Docker engine wouldn't start with the previous
 command

---
 airflow/docker/cwl/docker_cwl_entrypoint.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/airflow/docker/cwl/docker_cwl_entrypoint.sh b/airflow/docker/cwl/docker_cwl_entrypoint.sh
index da2aacc4..41916d0d 100755
--- a/airflow/docker/cwl/docker_cwl_entrypoint.sh
+++ b/airflow/docker/cwl/docker_cwl_entrypoint.sh
@@ -18,7 +18,7 @@ cat /tmp/job_args.json
 mkdir -p "$output_dir"
 
 # Start Docker engine
-dockerd > dockerd-logfile 2>&1
+dockerd &> dockerd-logfile &
 
 # Wait until Docker engine is running
 # Loop until 'docker version' exits with 0.