Generic Resource Object Web (GROW)

about

This is a simple proof of concept program.
It is a Go based program that serves objects. These objects come from an S3 API object store. It can use the open source Minio S3 backend or any of the AWS, Google Cloud or Azure object storage services. I've tested this mostly on Docker Swarm and Google Cloud Run.

Note: GROW is a very simplistic object server. For a more sophisticate solution with far more features try Clouder. (https://clowder.ncsa.illinois.edu/)

GROW leverages the RDA Digital Object Cloud pattern and is a basic implementation of that pattern.

For more details visit the about page.

commands

GROW can be run from the command line for Docker.

example command line with object store option

go run cmd/server/main.go -domain "https://example.org" -server 192.168.86.45:1234 
-bucket sites -prefix siteprefix -key mykey -secret mysecret

The elements are:

domain: Your sites domain (needed when generating sitemaps)

server: The address for the object storage such as Minio

bucket: The bucket your object tree is stored in

prefix:  Optional prefix for your object tree root

key:  Object store key

secret: Object store secret

On Docker this would look like:

    image: fils/grow-general:latest
    environment:
      - S3ADDRESS=s3system:9000
      - S3BUCKET=sites
      - S3PREFIX=siteprefix
      - DOMAIN=https://examples.org/
      - S3KEY=mykey
      - S3SECRET=mysecret
    labels:
    ...
    networks:
      - traefik_default

Future Thoughts

I am interested in packages like caffeine that represet a REST style approach. They could be used directly or as a pattern in GROW to test structured data on the web tools.