config(rest-api): enable authentication and api authorization by defa…

…ult (#11484)
datahub-project · Sep 27, 2024 · 4296373 · 4296373
1 parent 6a58493
commit 4296373
Show file tree

Hide file tree

Showing 76 changed files with 1,864 additions and 1,950 deletions.
diff --git a/datahub-frontend/run/frontend.env b/datahub-frontend/run/frontend.env
@@ -44,7 +44,7 @@ ELASTIC_CLIENT_PORT=9200
 # AUTH_JAAS_ENABLED=false
 
 # Change to disable Metadata Service Authentication
-METADATA_SERVICE_AUTH_ENABLED=true
+# METADATA_SERVICE_AUTH_ENABLED=false
 
 # Change to override max header count defaults
 DATAHUB_AKKA_MAX_HEADER_COUNT=64

diff --git a/docker/datahub-frontend/env/docker.env b/docker/datahub-frontend/env/docker.env
@@ -16,8 +16,8 @@ JAVA_OPTS=-Xms512m -Xmx512m -Dhttp.port=9002 -Dconfig.file=datahub-frontend/conf
 # SSL_TRUSTSTORE_TYPE=jks
 # SSL_TRUSTSTORE_PASSWORD=MyTruststorePassword
 
-# Uncomment to enable Metadata Service Authentication
-# METADATA_SERVICE_AUTH_ENABLED=true
+# Uncomment to disable Metadata Service Authentication
+# METADATA_SERVICE_AUTH_ENABLED=false
 
 # Uncomment & populate these configs to enable OIDC SSO in React application.
 # Required OIDC configs

diff --git a/docker/datahub-gms/env/docker.env b/docker/datahub-gms/env/docker.env
@@ -27,8 +27,8 @@ MCE_CONSUMER_ENABLED=true
 PE_CONSUMER_ENABLED=true
 UI_INGESTION_ENABLED=true
 
-# Uncomment to enable Metadata Service Authentication
-METADATA_SERVICE_AUTH_ENABLED=false
+# Uncomment to disable Metadata Service Authentication
+# METADATA_SERVICE_AUTH_ENABLED=false
 
 # Uncomment to disable persistence of client-side analytics events
 # DATAHUB_ANALYTICS_ENABLED=false

diff --git a/docker/docker-compose-with-cassandra.yml b/docker/docker-compose-with-cassandra.yml
@@ -39,6 +39,8 @@ services:
         context: ../
         dockerfile: docker/datahub-gms/Dockerfile
     env_file: ./datahub-gms/env/docker.cassandra.env
+    environment:
+      - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     healthcheck:
       test: curl -sS --fail http://datahub-gms:${DATAHUB_GMS_PORT:-8080}/health
       start_period: 20s

diff --git a/docker/docker-compose-without-neo4j.override.yml b/docker/docker-compose-without-neo4j.override.yml
@@ -6,6 +6,7 @@ services:
     environment:
       - DATAHUB_SERVER_TYPE=${DATAHUB_SERVER_TYPE:-quickstart}
       - DATAHUB_TELEMETRY_ENABLED=${DATAHUB_TELEMETRY_ENABLED:-true}
+      - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     depends_on:
       datahub-upgrade:
         condition: service_completed_successfully

diff --git a/docker/docker-compose-without-neo4j.postgres.override.yml b/docker/docker-compose-without-neo4j.postgres.override.yml
@@ -9,6 +9,7 @@ services:
     environment:
       - DATAHUB_SERVER_TYPE=${DATAHUB_SERVER_TYPE:-quickstart}
       - DATAHUB_TELEMETRY_ENABLED=${DATAHUB_TELEMETRY_ENABLED:-true}
+      - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     depends_on:
       datahub-upgrade:
         condition: service_completed_successfully

diff --git a/docker/docker-compose-without-neo4j.yml b/docker/docker-compose-without-neo4j.yml
@@ -42,6 +42,7 @@ services:
     env_file: datahub-gms/env/docker-without-neo4j.env
     environment:
       - KAFKA_CONSUMER_STOP_ON_DESERIALIZATION_ERROR=${KAFKA_CONSUMER_STOP_ON_DESERIALIZATION_ERROR:-true}
+      - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     healthcheck:
       test: curl -sS --fail http://datahub-gms:${DATAHUB_GMS_PORT:-8080}/health
       start_period: 90s

diff --git a/docker/docker-compose.dev.yml b/docker/docker-compose.dev.yml
@@ -40,7 +40,7 @@ services:
     - SKIP_ELASTICSEARCH_CHECK=false
     - DATAHUB_SERVER_TYPE=${DATAHUB_SERVER_TYPE:-dev}
     - DATAHUB_TELEMETRY_ENABLED=${DATAHUB_TELEMETRY_ENABLED:-true}
-    - METADATA_SERVICE_AUTH_ENABLED=false
+    - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     - JAVA_TOOL_OPTIONS=-agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=*:5001
     - BOOTSTRAP_SYSTEM_UPDATE_WAIT_FOR_SYSTEM_UPDATE=false
     - SEARCH_SERVICE_ENABLE_CACHE=false

diff --git a/docker/docker-compose.override.yml b/docker/docker-compose.override.yml
@@ -8,6 +8,7 @@ services:
     - DATAHUB_SERVER_TYPE=${DATAHUB_SERVER_TYPE:-quickstart}
     - DATAHUB_TELEMETRY_ENABLED=${DATAHUB_TELEMETRY_ENABLED:-true}
     - GRAPH_SERVICE_IMPL=${GRAPH_SERVICE_IMPL:-elasticsearch}
+    - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     volumes:
     - ${HOME}/.datahub/plugins:/etc/datahub/plugins
   datahub-upgrade:

diff --git a/docker/docker-compose.yml b/docker/docker-compose.yml
@@ -35,6 +35,7 @@ services:
     image: ${DATAHUB_GMS_IMAGE:-acryldata/datahub-gms}:${DATAHUB_VERSION:-head}
     environment:
       - KAFKA_CONSUMER_STOP_ON_DESERIALIZATION_ERROR=${KAFKA_CONSUMER_STOP_ON_DESERIALIZATION_ERROR:-true}
+      - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     ports:
     - ${DATAHUB_MAPPED_GMS_PORT:-8080}:8080
     build:

diff --git a/docker/profiles/docker-compose.gms.yml b/docker/profiles/docker-compose.gms.yml
@@ -101,6 +101,7 @@ x-datahub-gms-service: &datahub-gms-service
   environment: &datahub-gms-env
     <<: [*primary-datastore-mysql-env, *graph-datastore-search-env, *search-datastore-env, *datahub-quickstart-telemetry-env, *kafka-env]
     ELASTICSEARCH_QUERY_CUSTOM_CONFIG_FILE: ${ELASTICSEARCH_QUERY_CUSTOM_CONFIG_FILE:-search_config.yaml}
+    METADATA_SERVICE_AUTH_ENABLED: ${METADATA_SERVICE_AUTH_ENABLED:-true}
   healthcheck:
     test: curl -sS --fail http://datahub-gms:${DATAHUB_GMS_PORT:-8080}/health
     start_period: 90s

diff --git a/docker/quickstart/docker-compose-m1.quickstart.yml b/docker/quickstart/docker-compose-m1.quickstart.yml
@@ -97,7 +97,7 @@ services:
     - KAFKA_SCHEMAREGISTRY_URL=http://schema-registry:8081
     - MAE_CONSUMER_ENABLED=true
     - MCE_CONSUMER_ENABLED=true
-    - METADATA_SERVICE_AUTH_ENABLED=false
+    - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     - NEO4J_HOST=http://neo4j:7474
     - NEO4J_PASSWORD=datahub
     - NEO4J_URI=bolt://neo4j

diff --git a/docker/quickstart/docker-compose-without-neo4j-m1.quickstart.yml b/docker/quickstart/docker-compose-without-neo4j-m1.quickstart.yml
@@ -97,6 +97,7 @@ services:
     - KAFKA_SCHEMAREGISTRY_URL=http://schema-registry:8081
     - MAE_CONSUMER_ENABLED=true
     - MCE_CONSUMER_ENABLED=true
+    - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     - PE_CONSUMER_ENABLED=true
     - UI_INGESTION_ENABLED=true
     healthcheck:

diff --git a/docker/quickstart/docker-compose-without-neo4j.quickstart.yml b/docker/quickstart/docker-compose-without-neo4j.quickstart.yml
@@ -97,6 +97,7 @@ services:
     - KAFKA_SCHEMAREGISTRY_URL=http://schema-registry:8081
     - MAE_CONSUMER_ENABLED=true
     - MCE_CONSUMER_ENABLED=true
+    - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     - PE_CONSUMER_ENABLED=true
     - UI_INGESTION_ENABLED=true
     healthcheck:

diff --git a/docker/quickstart/docker-compose.quickstart.yml b/docker/quickstart/docker-compose.quickstart.yml
@@ -97,7 +97,7 @@ services:
     - KAFKA_SCHEMAREGISTRY_URL=http://schema-registry:8081
     - MAE_CONSUMER_ENABLED=true
     - MCE_CONSUMER_ENABLED=true
-    - METADATA_SERVICE_AUTH_ENABLED=false
+    - METADATA_SERVICE_AUTH_ENABLED=${METADATA_SERVICE_AUTH_ENABLED:-false}
     - NEO4J_HOST=http://neo4j:7474
     - NEO4J_PASSWORD=datahub
     - NEO4J_URI=bolt://neo4j

diff --git a/docs/how/updating-datahub.md b/docs/how/updating-datahub.md
@@ -20,6 +20,9 @@ This file documents any backwards-incompatible changes in DataHub and assists pe
 
 ### Breaking Changes
 
+- Metadata service authentication enabled by default
+- Rest API authorization enabled by default
+
 ### Potential Downtime
 
 ### Deprecations

diff --git a/metadata-ingestion/src/datahub/api/entities/structuredproperties/structuredproperties.py b/metadata-ingestion/src/datahub/api/entities/structuredproperties/structuredproperties.py
@@ -94,60 +94,59 @@ def urn_must_be_present(cls, v, values):
         return v
 
     @staticmethod
-    def create(file: str) -> None:
-        emitter: DataHubGraph
-
-        with get_default_graph() as emitter:
-            with open(file) as fp:
-                structuredproperties: List[dict] = yaml.safe_load(fp)
-                for structuredproperty_raw in structuredproperties:
-                    structuredproperty = StructuredProperties.parse_obj(
-                        structuredproperty_raw
+    def create(file: str, graph: Optional[DataHubGraph] = None) -> None:
+        emitter: DataHubGraph = graph if graph else get_default_graph()
+
+        with open(file) as fp:
+            structuredproperties: List[dict] = yaml.safe_load(fp)
+            for structuredproperty_raw in structuredproperties:
+                structuredproperty = StructuredProperties.parse_obj(
+                    structuredproperty_raw
+                )
+                if not structuredproperty.type.islower():
+                    structuredproperty.type = structuredproperty.type.lower()
+                    logger.warn(
+                        f"Structured property type should be lowercase. Updated to {structuredproperty.type}"
                     )
-                    if not structuredproperty.type.islower():
-                        structuredproperty.type = structuredproperty.type.lower()
-                        logger.warn(
-                            f"Structured property type should be lowercase. Updated to {structuredproperty.type}"
-                        )
-                    if not AllowedTypes.check_allowed_type(structuredproperty.type):
-                        raise ValueError(
-                            f"Type {structuredproperty.type} is not allowed. Allowed types are {AllowedTypes.values()}"
-                        )
-                    mcp = MetadataChangeProposalWrapper(
-                        entityUrn=structuredproperty.urn,
-                        aspect=StructuredPropertyDefinitionClass(
-                            qualifiedName=structuredproperty.fqn,
-                            valueType=Urn.make_data_type_urn(structuredproperty.type),
-                            displayName=structuredproperty.display_name,
-                            description=structuredproperty.description,
-                            entityTypes=[
-                                Urn.make_entity_type_urn(entity_type)
-                                for entity_type in structuredproperty.entity_types or []
-                            ],
-                            cardinality=structuredproperty.cardinality,
-                            immutable=structuredproperty.immutable,
-                            allowedValues=(
-                                [
-                                    PropertyValueClass(
-                                        value=v.value, description=v.description
-                                    )
-                                    for v in structuredproperty.allowed_values
-                                ]
-                                if structuredproperty.allowed_values
-                                else None
-                            ),
-                            typeQualifier=(
-                                {
-                                    "allowedTypes": structuredproperty.type_qualifier.allowed_types
-                                }
-                                if structuredproperty.type_qualifier
-                                else None
-                            ),
-                        ),
+                if not AllowedTypes.check_allowed_type(structuredproperty.type):
+                    raise ValueError(
+                        f"Type {structuredproperty.type} is not allowed. Allowed types are {AllowedTypes.values()}"
                     )
-                    emitter.emit_mcp(mcp)
+                mcp = MetadataChangeProposalWrapper(
+                    entityUrn=structuredproperty.urn,
+                    aspect=StructuredPropertyDefinitionClass(
+                        qualifiedName=structuredproperty.fqn,
+                        valueType=Urn.make_data_type_urn(structuredproperty.type),
+                        displayName=structuredproperty.display_name,
+                        description=structuredproperty.description,
+                        entityTypes=[
+                            Urn.make_entity_type_urn(entity_type)
+                            for entity_type in structuredproperty.entity_types or []
+                        ],
+                        cardinality=structuredproperty.cardinality,
+                        immutable=structuredproperty.immutable,
+                        allowedValues=(
+                            [
+                                PropertyValueClass(
+                                    value=v.value, description=v.description
+                                )
+                                for v in structuredproperty.allowed_values
+                            ]
+                            if structuredproperty.allowed_values
+                            else None
+                        ),
+                        typeQualifier=(
+                            {
+                                "allowedTypes": structuredproperty.type_qualifier.allowed_types
+                            }
+                            if structuredproperty.type_qualifier
+                            else None
+                        ),
+                    ),
+                )
+                emitter.emit_mcp(mcp)
 
-                    logger.info(f"Created structured property {structuredproperty.urn}")
+                logger.info(f"Created structured property {structuredproperty.urn}")
 
     @classmethod
     def from_datahub(cls, graph: DataHubGraph, urn: str) -> "StructuredProperties":

diff --git a/metadata-ingestion/src/datahub/cli/timeline_cli.py b/metadata-ingestion/src/datahub/cli/timeline_cli.py
@@ -9,7 +9,7 @@
 from termcolor import colored
 
 from datahub.emitter.mce_builder import dataset_urn_to_key, schema_field_urn_to_key
-from datahub.ingestion.graph.client import get_default_graph
+from datahub.ingestion.graph.client import DataHubGraph, get_default_graph
 from datahub.telemetry import telemetry
 from datahub.upgrade import upgrade
 from datahub.utilities.urns.urn import Urn
@@ -62,8 +62,9 @@ def get_timeline(
     start_time: Optional[int],
     end_time: Optional[int],
     diff: bool,
+    graph: Optional[DataHubGraph] = None,
 ) -> Any:
-    client = get_default_graph()
+    client = graph if graph else get_default_graph()
     session = client._session
     host = client.config.server
     if urn.startswith("urn%3A"):

diff --git a/...ce-consumer-job/src/test/java/com/linkedin/metadata/kafka/MceConsumerApplicationTest.java b/...ce-consumer-job/src/test/java/com/linkedin/metadata/kafka/MceConsumerApplicationTest.java
@@ -2,7 +2,8 @@
 
 import static org.mockito.ArgumentMatchers.any;
 import static org.mockito.Mockito.when;
-import static org.testng.AssertJUnit.*;
+import static org.testng.Assert.assertNotNull;
+import static org.testng.Assert.assertTrue;
 
 import com.linkedin.metadata.entity.EntityService;
 import com.linkedin.metadata.entity.restoreindices.RestoreIndicesResult;
@@ -19,7 +20,8 @@
 @ActiveProfiles("test")
 @SpringBootTest(
     webEnvironment = SpringBootTest.WebEnvironment.RANDOM_PORT,
-    classes = {MceConsumerApplication.class, MceConsumerApplicationTestConfiguration.class})
+    classes = {MceConsumerApplication.class, MceConsumerApplicationTestConfiguration.class},
+    properties = {"authentication.enabled=false"})
 public class MceConsumerApplicationTest extends AbstractTestNGSpringContextTests {
 
   @Autowired private TestRestTemplate restTemplate;
@@ -38,7 +40,7 @@ public void testRestliServletConfig() {
     String response =
         this.restTemplate.postForObject(
             "/gms/aspects?action=restoreIndices", "{\"urn\":\"\"}", String.class);
-    assertTrue(response.contains(mockResult.toString()));
+    assertTrue(response.contains(mockResult.toString()), String.format("Found: %s", response));
   }
 
   @Test

diff --git a/metadata-service/configuration/src/main/resources/application.yaml b/metadata-service/configuration/src/main/resources/application.yaml
@@ -3,8 +3,8 @@ baseUrl: ${DATAHUB_BASE_URL:http://localhost:9002}
 
 # App Layer
 authentication:
-  # Enable if you want all requests to the Metadata Service to be authenticated. Disabled by default.
-  enabled: ${METADATA_SERVICE_AUTH_ENABLED:false}
+  # Enable if you want all requests to the Metadata Service to be authenticated.
+  enabled: ${METADATA_SERVICE_AUTH_ENABLED:true}
 
   # Required if enabled is true! A configurable chain of Authenticators
   authenticators:
@@ -43,8 +43,8 @@ authorization:
     enabled: ${AUTH_POLICIES_ENABLED:true}
     cacheRefreshIntervalSecs: ${POLICY_CACHE_REFRESH_INTERVAL_SECONDS:120}
     cachePolicyFetchSize: ${POLICY_CACHE_FETCH_SIZE:1000}
-  # Enables authorization of reads, writes, and deletes on REST APIs. Defaults to false for backwards compatibility, but should become true down the road
-  restApiAuthorization: ${REST_API_AUTHORIZATION_ENABLED:false}
+  # Enables authorization of reads, writes, and deletes on REST APIs.
+  restApiAuthorization: ${REST_API_AUTHORIZATION_ENABLED:true}
   view:
     enabled: ${VIEW_AUTHORIZATION_ENABLED:false}
     recommendations:

diff --git a/...api-servlet/src/main/java/io/datahubproject/openapi/GlobalControllerExceptionHandler.java b/...api-servlet/src/main/java/io/datahubproject/openapi/GlobalControllerExceptionHandler.java
@@ -2,6 +2,7 @@
 
 import com.linkedin.metadata.dao.throttle.APIThrottleException;
 import io.datahubproject.openapi.exception.InvalidUrnException;
+import io.datahubproject.openapi.exception.UnauthorizedException;
 import java.util.Map;
 import lombok.extern.slf4j.Slf4j;
 import org.springframework.beans.ConversionNotSupportedException;
@@ -45,4 +46,10 @@ public static ResponseEntity<Map<String, String>> handleThrottleException(
     return new ResponseEntity<>(
         Map.of("error", e.getMessage()), headers, HttpStatus.TOO_MANY_REQUESTS);
   }
+
+  @ExceptionHandler(UnauthorizedException.class)
+  public static ResponseEntity<Map<String, String>> handleUnauthorizedException(
+      UnauthorizedException e) {
+    return new ResponseEntity<>(Map.of("error", e.getMessage()), HttpStatus.FORBIDDEN);
+  }
 }
diff --git a/smoke-test/conftest.py b/smoke-test/conftest.py
@@ -0,0 +1,39 @@
+import os
+
+import pytest
+import requests
+from datahub.ingestion.graph.client import DatahubClientConfig, DataHubGraph
+
+from tests.test_result_msg import send_message
+from tests.utils import (
+    TestSessionWrapper,
+    get_frontend_session,
+    wait_for_healthcheck_util,
+)
+
+# Disable telemetry
+os.environ["DATAHUB_TELEMETRY_ENABLED"] = "false"
+
+
+@pytest.fixture(scope="session")
+def auth_session():
+    wait_for_healthcheck_util(requests)
+    auth_session = TestSessionWrapper(get_frontend_session())
+    yield auth_session
+    auth_session.destroy()
+
+
+@pytest.fixture(scope="session")
+def graph_client(auth_session) -> DataHubGraph:
+    print(auth_session.cookies)
+    graph: DataHubGraph = DataHubGraph(
+        config=DatahubClientConfig(
+            server=auth_session.gms_url(), token=auth_session.gms_token()
+        )
+    )
+    return graph
+
+
+def pytest_sessionfinish(session, exitstatus):
+    """whole test run finishes."""
+    send_message(exitstatus)
diff --git a/smoke-test/requests_wrapper/__init__.py b/smoke-test/requests_wrapper/__init__.py
diff --git a/smoke-test/requests_wrapper/constants.py b/smoke-test/requests_wrapper/constants.py