This repository servers two purposes.
- It contains the manifest of which versions of which components are deployed to each environment.
- It is the home of system-wide integration tests.
Before you can run the tests, you need to setup your environment:
pip install -r requirements.txt
# ping in #dcp-ops or #dcp-ops-help if you need valid service account credentials
export GOOGLE_APPLICATION_CREDENTIALS={PATH_TO_VALID_SERVICE_ACCOUNT_CREDS_WITH_ACCESS_TO_DSS}
Instructions for running tests locally in python 3.6:
Be aware that you will be uploading bundles to the specified deployment (integration/staging/prod) of the DCP and running secondary analysis. The test currently takes about 120 minutes to run (longer for optimus, shorter for smart-seq-2).
DEPLOYMENT_ENV={ENV} make
To only run one of the pipelines, smart-seq2 for example, and retain the bundles (disable tombstoning) use Python 3.6:
RETAIN_BUNDLES=True DEPLOYMENT_ENV={ENV} python -m unittest tests.integration.test_end_to_end_dcp.TestSmartSeq2Run.test_smartseq2_run
To only run one of the pipelines, smart-seq2 for example, use Python 3.6:
DEPLOYMENT_ENV={ENV} python -m unittest tests.integration.test_end_to_end_dcp.TestSmartSeq2Run.test_smartseq2_run
See our Security Policy.
See https://allspark.dev.data.humancellatlas.org/dcp-ops/docs/wikis/home.