Open ddalton-swe opened 7 months ago
There were issues encountered when running this within MAAP's ADE. The issues are related to running Docker within Docker. This issue will be debugged with some of the MAAP developers
Future tasks.
If the run is successful then the next step would be to register IMI as an algorithm within MAAP. There should also be a report generated of lessons learned and issues that are found when running the algorithm. Run forward and inverse models of IMI
@laestrada are you able to provide some of the files that are listed in the config.harvard-cannon.txt config file? Mainly things like
StateVectorFile: "/n/holyscratch01/jacob_lab/msulprizio/Test_IMI/StateVector.nc"
ShapeFile: "resources/shapefiles/PermianBasin_Extent_201712.shp"
Hi @ddalton-jpl, that file is only needed if CreateAutomaticRectilinearStateVectorFile
is true
. Our benchmark case builds its own state vector, so you shouldn't need the custom shapefile or statevector
@laestrada What about for spin-up?
src/components/spinup_component/spinup.sh: line 89: cd: /n/holyscratch01/jacob_lab//Test_IMI/spinup_run: No such file or directory
Or is that even needed?
Note:
@laestrada and I got this to run within MAAP via Docker. There are some issues with MAAP disconnecting the workspace which causes all packages to be removed. The credential issue was resolved by providing the aws_session_token to an environments.env file. The issue with Docker was resolved by changing the workspace DevFile to have privledged=true
The goal for this task is to run the containerized IMI within the MAAP ADE.