We run an assortment of integration tests flows for the purpose of evaluating the health of our systems. Defined below are the locations in which certain flows run.
We run several flows on a regular interval in both the dev & stg environments. Unexpected failures are reported to the relevant environment channel #alerts-ENV-cloud2.
We utilize two work queues for test flows in our production environment:
- integration-tests
- helm
The integration-tests work queue runs our typical test flows.
The helm work queue runs a few simple flows against the latest version of a prefect worker. We automatically deploy the latest helm chart version upon it's release to this environement. This helps to ensure that the worker chart is always in a working state.
Similarly to the helm work queue, we run the latest prefect server & worker, automatically updated upon release. We run a similar set of test flows here to ensure the latest release of the server chart is is always in a working state.
To re-register all of these flows, first adjust your prefect profile to contain the following:
[profiles.server]
PREFECT_API_URL = "http://server.private.prefect.cloud/api"
Then run:
prefect profiles use server
prefect --no-prompt deploy --name "Flow Pauses" --name "Flow Results" --name "Flow Retries With Subflows" --name "Flow Retries" --name "Hello Tasks" --name "Secret Block" --name "Task Burst" --name "Task Results" --name "Task Retries"