Deployable Science Service using Jupyter and CWL
Scripts run veriations of the following too commands:
docker build t sci-serv docker/
docker run -it -p 8888:8888 sci-serv
The docker commands are packaged in:
bash build-docker.sh
To run a docker instance use:
bash run-docker.sh
The container is accessable using a jupyter notebook. Also available as a datascience enabled instance, with an alternate script.
Copy/paste this URL into your browser when you connect for the first time, to login with a token: http://localhost:8888/?token=6a287b208b22105f161bfaea8ab8b4725b0174b73d8cd343
In Jupyter run the following script to see CWL running in the container.
bash science-service/run_workflow.sh
The container can be further staged for use in EPOS with an execution of a cwl input at run time:
./exec-docker.sh sci_serv_container_name inputs.yml
Note: To run the container service via the exec script requires a valid container instance and a input file containing multiple OGC Service links.
#mihajlov@pc150396 science-service]$ docker images
| REPOSITORY | TAG | IMAGE ID | SIZE |
|---|---|---|---|
| sci-serv | latest | 7b1461198760 | 6.34 GB |
| docker.io/jupyter/datascience-notebook | latest | db148b62d457 | 6.25 GB |
| docker.io/jupyter/minimal-notebook | latest | c4c8a0c902ee | 2.72 GB |
| docker.io/ubuntu | 14.04 | 67759a80360c | 221 MB |
| docker.io/commonworkflowlanguage/cwltool_module | latest | 1a55ee22efea | 81 MB |