Developing your own app using Open OnDemand

We have a set of applications based on software usage/user requests added to Open OnDemand but OnDemand provide a way to develop or deploy the existing application developed by the OnDemand team like here or you find on GitHub like here

You will need to follow the steps below to enable the development icon in your top bar (menu) of OnDemand.

To enable your development environment you should:

1. connect to FAS-OnDemand for Cannon or FASSE

2.  open a shell from the OnDemand or connect to login nodes and then create the development folder as mkdir $HOME/.fasrcood/dev . On FASSE cluster, the folder is $HOME/.fasseood/dev.

3. Now restart the webserver going to OnDemand.
4. Now you can access your sandbox development environment, either clicking on the item in the Develop menu in the dashboard, or simply navigating from the shell to $HOME/<OnDemand_folder>/dev and create an app folder in there and start your development.                                     
5. A simple way to test is to check out an app from https://osc.github.io/ood-documentation/release-2.0/install-ihpc-apps.html and update the module names/versions and launch the application.       
6. We do have Jupyter or RStudio for Academic cluster so you can git clone these into the dev folder as below and restart the webserver and go to My Sandbox Apps to access.
7.  To make the application work, you will need to update few files. I am listing few below as you will need to update the settings

https://github.com/fasrc/fas-ondemand-rstudio/blob/master/form.yml.erb

Update cluster: “academic” to “odyssey3” or “fasse” based on cluster and bc_queue: “academic” to “shared” or “fasse”.  You may need to update number of cores also based on the cluster and partition

https://github.com/fasrc/fas-ondemand-rstudio/blob/master/template/script.sh.erb

This file has the location for the Singularity container and this location is not accessible on the Cannon or FASSE cluster so you need to update the location of the container based on the container you have created.

Feel free to reach out to rchelp@rc.fas.harvard.edu or visit our office hours to help deploy a container or answer further questions.