opendatahub-io-contrib / airflow-on-openshift Goto Github PK

The application is currently not serving requests at this endpoint. It may not have been started or is still starting.

Post deployment, I see the route, but I do not see any other flower resources or services that are being exposed to route. In other deployments, I see the configs for the flower service defined in the airflow-values.yaml (https://github.com/redhat-na-ssa/mlops-prototype/blob/main/airflow/airflow-setup/airflow-values.yaml)

A user can use the official airflow helm chart to allow ingress with k8s ingress objects - this requires defining the host.

We want to document this as another method of setting up ingress using the official helm chart.

The current instructions point to this project and dags folder, where I would not host my dags
--set dags.gitSync.repo=https://github.com/opendatahub-io-contrib/airflow-on-openshift.git
--set dags.gitSync.branch=main
--set dags.gitSync.subPath=dags

Propose variables or documentation on how to point the deployment to the users and
--set dags.gitSync.repo=https://github.com/<YOUR_NAME>/<YOUR_REPOSITORY>.git
--set dags.gitSync.branch=main
--set dags.gitSync.subPath=<YOUR_PATH>/<YOUR_DAGS_DIRECTORY>

We want to explore various options around postgres deployment in combination with airflow.

Some options today are:

• OpenShift Template
• Crunchy Postgres Operator
• Airflow Helm Values

Explore using the stackable operator for airflow

https://github.com/stackabletech/airflow-operator

Clean up script to remove resources deployed

Installed the operator through the helm chart according to https://github.com/opendatahub-io-contrib/airflow-on-openshift#install-airflow-via---set on an RHPDS demo environment of OpenShift 4.12
Some pods start running properly but most don't:

All have the same error:

sqlalchemy.exc.OperationalError: (psycopg2.OperationalError) connection to server at "airflow-postgresql.airflow" (172.30.137.62), port 5432 failed: Connection refused
Is the server running on that host and accepting TCP/IP connections?

Hi there, I have installed airflow using the hacks/easy_install.sh and everything has been working fine. I am now trying to find a way to install additional python packages and I'm wondering if there's a recommended way to do this here? like, if a custom dockerfile with a RUN pip install kafka-python (e.g.) is the way to go or should another method be used? Sincerely appreciate any feedback