@@ -49,8 +49,8 @@ According to the [DAG implementation details](#dag-implementation-details) need
#### Installing Python Dependencies
Environment dependencies might be installed by several ways:
1.Setting up an environment into the Cloud Composer Console.
2.Installing local Python library. Put your dependencies into *DAG_FOLDER/libs* directory. Airflow automatically adds *DAG_FOLDER* and *PLUGINS_FOLDER* to the *PATH*.
1.Install packages via `pip` on the environment where Airflow runs.
2.Although it is **not recommended**, it is possible to install Python libraries locally. Put your dependencies into *DAG_FOLDER/libs* directory. Airflow automatically adds *DAG_FOLDER* and *PLUGINS_FOLDER* to the *PATH*.
### Azure
To deploy the Ingestion DAGs to airflow, follow below steps.
...
...
@@ -62,6 +62,13 @@ To deploy the Ingestion DAGs to airflow, follow below steps.
#### Installing Python Dependencies
Python dependencies can be specified as extra pip packages in airflow deployment [here](https://community.opengroup.org/osdu/platform/deployment-and-operations/infra-azure-provisioning/-/blob/master/charts/airflow/helm-config.yaml#L211)
Also, the DAGs require [Python SDK](https://community.opengroup.org/osdu/platform/system/sdks/common-python-sdk) to be installed.
It can be installed ton the environment via `pip`:
Add variables manually in the Airflow UI or through airflow helm charts. [List of the required variables](#required-variables).
...
...
@@ -74,7 +81,7 @@ More details about airflow variables can be found [here](https://airflow.apache.
OSDU DAGs are cloud platform-agnostic by design. However, there are specific implementation requirements by cloud
platforms, and the OSDU R2 Prototype provides a dedicated Python SDK to make sure that DAGs are independent from the
cloud platforms. This Python SDK is located in a separate [os-python-sdk] folder.
cloud platforms. This Python SDK must be installed from the corresponding [repository](https://community.opengroup.org/osdu/platform/system/sdks/common-python-sdk).