S2 ARD Processor

Docker container that runs the s2-ard-processor workflow.

The mapped input folder contains a set of S2 granules that will be processed as a swath. The processing can take place sequentially or in parallel using MPI on the JASMIN cluster.

Build and run instructions

Build to image:

cd workflow
docker build -t s2-ard-processor .

Use --no-cache to build from scratch

Run Interactively:

docker run -i --entrypoint /bin/bash 
    -v /<hostPath>/input:/input 
    -v /<hostPath>/output:/output 
    -v /<hostPath>/state:/state 
    -v /<hostPath>/static:/static 
    -v /<hostPath>/working:/working 
    -v /<hostPath>/report:/report
    -v /<hostPath>/database:/database 
    -t s2-ard-processor

Where <hostpath> is the path on the host to the mounted folder

Convert Docker image to apptainer image

Build an apptainer image using your Docker image

sudo apptainer build s2-ard-processor.sif docker-daemon://s2-ard-processor:latest

Run:

apptainer exec 
    --bind /<hostPath>/input:/input 
    --bind /<hostPath>/output:/output 
    --bind /<hostPath>/state:/state 
    --bind /<hostPath>/static:/static 
    --bind /<hostPath>/working:/working
    --bind /<hostPath>/report:/report
    --bind /<hostPath>/database:/database
    
    s2-ard-processor.sif /app/exec.sh 
        GenerateReport
        --dem=dem.kea 
        --outWkt=outwkt.txt 
        --projAbbv=osgb
        --metadataConfigFile=metadata.config.json 
        --metadataTemplate=metadataTemplate.xml
        --reportFileName=reportfile.csv
        --dbFileName=s2ardProducts.db
        --local-scheduler

For the full list of parameters and more details on the folder setups, see the workflow readme at workflow/app/workflows/README.md.

Code change and deployment process

The code in this repo will be jointly maintained by JNCC and DEFRA/CGI. Use the steps below as a guideline for making new changes:

1. Create a new feature branch from main and commit your changes there until you're ready to merge
2. Open a pull request to merge back into main and add a reviewer from both JNCC and CGI to notify them
3. JNCC then uses the feature branch to build a jncc/s2-ard-processor-dev docker image which both parties can use for testing and QA
4. Once it passes QA, JNCC will approve the PR, merge it into main, and build a live jncc/s2-ard-processor docker image which can be deployed to production