Skip to main content
Version: Next

Custom Properties

Why Would You Use Custom Properties on Datasets?​

Custom properties to datasets can help to provide additional information about the data that is not readily available in the standard metadata fields. Custom properties can be used to describe specific attributes of the data, such as the units of measurement used, the date range covered, or the geographical region the data pertains to. This can be particularly helpful when working with large and complex datasets, where additional context can help to ensure that the data is being used correctly and effectively.

DataHub models custom properties of a Dataset as a map of key-value pairs of strings.

Custom properties can also be used to enable advanced search and discovery capabilities, by allowing users to filter and sort datasets based on specific attributes. This can help users to quickly find and access the data they need, without having to manually review large numbers of datasets.

Goal Of This Guide​

This guide will show you how to add, remove or replace custom properties on a dataset fct_users_deleted. Here's what each operation means:

  • Add: Add custom properties to a dataset without affecting existing properties
  • Remove: Removing specific properties from the dataset without affecting other properties
  • Replace: Completely replacing the entire property map without affecting other fields that are located in the same aspect. e.g. DatasetProperties aspect contains customProperties as well as other fields like name and description.

Prerequisites​

For this tutorial, you need to deploy DataHub Quickstart and ingest sample data. For detailed information, please refer to Datahub Quickstart Guide.

note

Before adding custom properties, you need to ensure the target dataset is already present in your DataHub instance. If you attempt to manipulate entities that do not exist, your operation will fail. In this guide, we will be using data from sample ingestion.

In this example, we will add some custom properties cluster_name and retention_time to the dataset fct_users_deleted.

After you have ingested sample data, the dataset fct_users_deleted should have a custom properties section with encoding set to utf-8.

datahub get --urn "urn:li:dataset:(urn:li:dataPlatform:hive,fct_users_deleted,PROD)" --aspect datasetProperties
{
  "datasetProperties": {
    "customProperties": {
      "encoding": "utf-8"
    },
    "description": "table containing all the users deleted on a single day",
    "tags": []
  }
}

Add Custom Properties programmatically​

The following code adds custom properties cluster_name and retention_time to a dataset named fct_users_deleted without affecting existing properties.

# Inlined from /metadata-ingestion/examples/library/dataset_add_custom_properties_patch.py
from datahub.emitter.mce_builder import make_dataset_urn
from datahub.ingestion.graph.client import DataHubGraph, DataHubGraphConfig
from datahub.specific.dataset import DatasetPatchBuilder

# Create DataHub Client
datahub_client = DataHubGraph(DataHubGraphConfig(server="http://localhost:8080"))

# Create Dataset URN
dataset_urn = make_dataset_urn(platform="hive", name="fct_users_created", env="PROD")

# Create Dataset Patch to Add Custom Properties
patch_builder = DatasetPatchBuilder(dataset_urn)
patch_builder.add_custom_property("cluster_name", "datahubproject.acryl.io")
patch_builder.add_custom_property("retention_time", "2 years")
patch_mcps = patch_builder.build()

# Emit Dataset Patch
for patch_mcp in patch_mcps:
    datahub_client.emit(patch_mcp)

Expected Outcome of Adding Custom Properties​

You can now see the two new properties are added to fct_users_deleted and the previous property encoding is unchanged.

We can also verify this operation by programmatically checking the datasetProperties aspect after running this code using the datahub cli.

datahub get --urn "urn:li:dataset:(urn:li:dataPlatform:hive,fct_users_deleted,PROD)" --aspect datasetProperties
{
  "datasetProperties": {
    "customProperties": {
      "encoding": "utf-8",
      "cluster_name": "datahubproject.acryl.io",
      "retention_time": "2 years"
    },
    "description": "table containing all the users deleted on a single day",
    "tags": []
  }
}

Add and Remove Custom Properties programmatically​

The following code shows you how can add and remove custom properties in the same call. In the following code, we add custom property cluster_name and remove property retention_time from a dataset named fct_users_deleted without affecting existing properties.

# Inlined from /metadata-ingestion/examples/library/dataset_add_remove_custom_properties_patch.py
from datahub.emitter.mce_builder import make_dataset_urn
from datahub.ingestion.graph.client import DataHubGraph, DataHubGraphConfig
from datahub.specific.dataset import DatasetPatchBuilder

# Create DataHub Client
datahub_client = DataHubGraph(DataHubGraphConfig(server="http://localhost:8080"))

# Create Dataset URN
dataset_urn = make_dataset_urn(platform="hive", name="fct_users_created", env="PROD")

# Create Dataset Patch to Add + Remove Custom Properties
patch_builder = DatasetPatchBuilder(dataset_urn)
patch_builder.add_custom_property("cluster_name", "datahubproject.acryl.io")
patch_builder.remove_custom_property("retention_time")
patch_mcps = patch_builder.build()

# Emit Dataset Patch
for patch_mcp in patch_mcps:
    datahub_client.emit(patch_mcp)

Expected Outcome of Add and Remove Operations on Custom Properties​

You can now see the cluster_name property is added to fct_users_deleted and the retention_time property is removed.

We can also verify this operation programmatically by checking the datasetProperties aspect using the datahub cli.

datahub get --urn "urn:li:dataset:(urn:li:dataPlatform:hive,fct_users_deleted,PROD)" --aspect datasetProperties
{
  "datasetProperties": {
    "customProperties": {
      "encoding": "utf-8",
      "cluster_name": "datahubproject.acryl.io"
    },
    "description": "table containing all the users deleted on a single day",
    "tags": []
  }
}

Replace Custom Properties programmatically​

The following code replaces the current custom properties with a new properties map that includes only the properties cluster_name and retention_time. After running this code, the previous encoding property will be removed.

# Inlined from /metadata-ingestion/examples/library/dataset_replace_properties.py
import logging
from typing import Union

from datahub.configuration.kafka import KafkaProducerConnectionConfig
from datahub.emitter.kafka_emitter import DatahubKafkaEmitter, KafkaEmitterConfig
from datahub.emitter.mce_builder import make_dataset_urn
from datahub.emitter.rest_emitter import DataHubRestEmitter
from datahub.specific.dataset import DatasetPatchBuilder

log = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO)


# Get an emitter, either REST or Kafka, this example shows you both
def get_emitter() -> Union[DataHubRestEmitter, DatahubKafkaEmitter]:
    USE_REST_EMITTER = True
    if USE_REST_EMITTER:
        gms_endpoint = "http://localhost:8080"
        return DataHubRestEmitter(gms_server=gms_endpoint)
    else:
        kafka_server = "localhost:9092"
        schema_registry_url = "http://localhost:8081"
        return DatahubKafkaEmitter(
            config=KafkaEmitterConfig(
                connection=KafkaProducerConnectionConfig(
                    bootstrap=kafka_server, schema_registry_url=schema_registry_url
                )
            )
        )


dataset_urn = make_dataset_urn(platform="hive", name="fct_users_created", env="PROD")

property_map_to_set = {
    "cluster_name": "datahubproject.acryl.io",
    "retention_time": "2 years",
}

with get_emitter() as emitter:
    for patch_mcp in (
        DatasetPatchBuilder(dataset_urn)
        .set_custom_properties(property_map_to_set)
        .build()
    ):
        emitter.emit(patch_mcp)


log.info(
    f"Replaced custom properties on dataset {dataset_urn} as {property_map_to_set}"
)

Expected Outcome of Replacing Custom Properties​

You can now see the cluster_name and retention_time properties are added to fct_users_deleted but the previous encoding property is no longer present.

We can also verify this operation programmatically by checking the datasetProperties aspect using the datahub cli.

datahub get --urn "urn:li:dataset:(urn:li:dataPlatform:hive,fct_users_deleted,PROD)" --aspect datasetProperties
{
  "datasetProperties": {
    "customProperties": {
      "cluster_name": "datahubproject.acryl.io",
      "retention_time": "2 years"
    },
    "description": "table containing all the users deleted on a single day",
    "tags": []
  }
}