Since 2018 we have been operating cloud-based JupyterHubs and BinderHubs for the community. These services were originally just experimental prototypes, designed to push forward the technology for cloud-native big-data geoscience. However, they rapidly gained significant numbers of users for day-to-day actual science. The most widely used example is ocean.pangeo.io, which runs on Google Cloud and has about 50 active daily users. These services have been funded by a direct grant of $100K worth of GCP credits from Google, via NSF’s BIGDATA program. Those funds have now run out, and we need to transition to a new mode of operation. Beyond the raw cost of the compute, there is of course also major human effort required to keep these services operating. Unfortunately, we cannot continue to provide free cloud computing to the world without funding.
Short Term Plan
Some ocean.pangeo.io users will be able to continue to use the service going forward, thanks to support from the Moore foundation. However, this will depend on the alignment of the research with the aims and scope of that funding.
Important: All users who wish to keep using ocean.pangeo.io must fill out this form by Monday, June 22:
ocean.pangeo.io will go down for maintenance on Monday, June 22 and will hopefully come back up later that week with new. Approved users will have their accounts and home directories migrated. Usernames will move from ORCID to GitHub user IDs. Current users who don’t fill out the form will have their accounts deleted.
All users should back up the data from their home directories. Here is an example of how to find all your notebooks and zip them in a tar archive for download.
find . -type f -name '*.ipynb' -print | grep -v '.ipynb_check' > list_of_notebooks tar -cvf backup.tar -T list_of_notebooks gzip backup.tar # download the backup.tar using jupyterlab interface
Long Term Plan
We firmly believe that Pangeo-style cloud computing has the potential to transform scientific research, making it more reproducible, transparent, and efficient. We are working hard at finding a sustainable long-term funding model that will permit all scientists to use the cloud in this way. This involves conversations with funders, cloud providers, and universities. To users inconvenienced by this transition, we are deeply sorry. Please stick with us as we prepare for the next phase. We welcome ideas and suggestions from the community on how to best most forward. Feel free to reply below with your thoughts or send me a DM.