cloudrun-recipe-handler

Pangeo Forge recipe handler for GCP Cloud Run.

To deploy, from within the /src directory:

$ gcloud run deploy $SERVICE_NAME --source .

To get service url from service name:

$ gcloud run services describe $SERVICE_NAME

To invoke:

$ curl \
  -X POST \
  -H "Authorization: Bearer $(gcloud auth print-identity-token)" \
  $SERVICE_URL \
  --json $PAYLOAD_JSON

or

$ curl \
  -X POST \
  -H "Authorization: Bearer $(gcloud auth print-identity-token)" \
  -H "Content-Type: application/json" \
  $SERVICE_URL \
  -d $PAYLOAD_JSON

Testing

pytest

Create and activate a local conda environment from the file ci/env.yaml:

$ conda env create --file ci/env.yaml
$ conda activate cloudrun

Then run the tests:

$ pytest src/tests

Note: Because one of the features of this service is to modify its own environment, the tests also modify the test environment. This means the tests are somewhat long-running, but in exchange for this performance cost, we get the reassurance of testing the actual behavior (as opposed to faster, but potentially inaccurate, mocked tests).

docker

The official Cloud Run docs have some good suggestions for testing locally with docker.