Install with Docker

Deploy

docker pull abarth/divand_rest
docker run --detach --name=divand_rest_container -it -p 8002:8002 abarth/divand_rest

Run docker rm divand_rest_container if the container already exists.

Testing

The tool jq needs to be installed to parse JSON from the shell. In Debian/Ubuntu, it can be installed by:

sudo apt-get install jq

The tests can be run with:

git clone https://github.com/gher-ulg/DIVAnd-REST.git
cd test
./test_bathymetry_curl.sh
./test_curl.sh # takes some time (~ minutes)

The analysis takes about 4 GB of RAM with the sample data. The first excution takes about 3 minutes and following requests about 1 minute for the sample data.

Open http://localhost:8002/ in your web browser, for the web user interface.

Debugging

Inspect logs with:

docker logs divand_rest_container

Developpement

*. start Julia and the DIVAnd server with include(joinpath("src","DIVAndREST.jl")) from the source directory *. start nginx:

/usr/sbin/nginx -c /home/DIVAnd/DIVAnd-REST/utils/nginx.conf

• run the test script

Open with with username and password (see note.md)

http://localhost:8002/?u=...&p=...

Containers deployed at CSC

Web user-interface:

https://diva.seadatacloud.ml/gui/1234/

Jupyterhub-interface:

https://diva.seadatacloud.ml/1234/tree?