can't access data from outside AWS & us-west-2

[gaelforget]

I was trying to run the tutorials from AGU workshop & 2021 Hackathon using mybinder.org yesterday.

Seems like it could be a good approach to let more people take advantage of the great stuff that you all have been putting together, outside of the time frame of specific events like AGU or the Hackathon. Any thoughts?

This worked to a point (you can try it at https://mybinder.org/v2/gh/gaelforget/corn/gfdev01) but what I mostly wanted to try out were the direct data access notebooks and these failed on mybinder (despite, I believe, having added my .netrc file in the right place). After looking more carefully at the repo webpage, my guess is that the errors I ran into had to do with direct data access being restricted to "AWS / us-west-2" and this brings a few questions :

• is my guess correct or am I overlooking another issue?
• is there or will there be a work around for this?
• would extending the S3 buckets to other regions / worldwide suffice ? (assuming that e.g. NASA would pay for that)

[betolink]

Hi @gaelforget your guess is correct, direct access (that is S3) is limited to us-west-2 and Binder does not get deployed there. There are workarounds. There are some free Jupyterhub deployments on us-west-2 you could use, i.e. Openscapes itself or Pangeo and some other institutions have them(upon previous authorization). Or, you could deploy your own EC2 instance and install Jupyterlab on it.

It's worth noticing that any cloud based dataset is also accessible outside AWS via a regular HTTPS link. e.g. s3://my_granule.h5 should also have a correspondent https://my_granule.h5 that's not limited to AWS regions.