How to integrate great-expectations with databricks and then to datahub

GX OSS Support
, , ,
1

I have to validate my data that is present on databricks unity catalog and then have to send the validation result to datahub.

I am trying this:

context_root_dir = "/dbfs/gx/"
context = gx.get_context(context_root_dir=context_root_dir)

dataframe_datasource = context.sources.add_or_update_spark(
    name="my_spark_in_memory_datasource"
)
df = spark.read.table("ecomm_checkout_catalog.checkout_schema.checkout_orders_data")
dataframe_asset = dataframe_datasource.add_dataframe_asset(
    name="checkout_schema.checkout_orders_data",
    dataframe= df,
)

batch_request1 = dataframe_asset.build_batch_request(dataframe=df)

expectation_suite_name = "my_data_expectations_suite"
context.add_or_update_expectation_suite(expectation_suite_name=expectation_suite_name)
validator = context.get_validator(
    batch_request=batch_request1,
    expectation_suite_name=expectation_suite_name,
)

validator.expect_column_values_to_not_be_null(column="orderid")
validator.save_expectation_suite(discard_failed_expectations=False)

my_checkpoint_name = "my_databricks_checkpoint"

checkpoint = Checkpoint(
    name=my_checkpoint_name,
    run_name_template="%Y%m%d-%H%M%S-my-run-name-template",
    data_context=context,
    batch_request=batch_request1,
    expectation_suite_name=expectation_suite_name,
    action_list=[
        {
            "name": "store_validation_result",
            "action": {
                "class_name": "StoreValidationResultAction"
            }
        },
        {
            "name": "update_data_docs",
            "action": {
                "class_name": "UpdateDataDocsAction"
            }
        },
        {
            "name": "datahub_action",
            "action": {
                "module_name": "datahub.integrations.great_expectations.action",
                "class_name": "DataHubValidationAction",
                "server_url": "http://35.240.254.***:8080"
            }
        }
    ]
)

context.add_or_update_checkpoint(checkpoint=checkpoint)
checkpoint_result = checkpoint.run()

Error:

DataHubValidationAction does not recognize this GE data asset type - <class 'great_expectations.validator.validator.Validator'>. This is either using v2-api or execution engine other than sqlalchemy.
Metadata not sent to datahub. No datasets found.
2

Hi @sant-singh, thanks for the question! Would you be able to provide what version of GX you’re using along with the rest of your stack?

3

Hi @nevintan , Yes, I am using latest version of GX
pip install great-expectations using this

4

It looks like DataHubValidationAction is set up to only use SqlAlchemyExecutionEngine. Have you configured your Data Context accordingly?

5

Datahub only support SqlAlchemyExecutionEngine
https://datahubproject.io/docs/metadata-ingestion/integration_docs/great-expectations/
Note: @Sant Narayan Singh originally posted this reply in Slack. It might not have transferred perfectly.

6

Can you provide a datasource congiguration with asset using SqlAlchemyExecutionEngine?

7

Have you taken a look at the attached loom video provided in the Datahub documentation?

8

Yes, I have seen that and It works well with postgresql but I have my delta tables in Databricks Unity Catalog(not in Databricks SQL warehouse). I am able to fetched that data and generated validation result but I am not able to send that validation result to datahub.