openmetadata-ingestion 1.9.13.0__py3-none-any.whl → 1.9.14.0__py3-none-any.whl

metadata/ingestion/source/pipeline/databrickspipeline/metadata.py

@@ -28,6 +28,7 @@ from metadata.generated.schema.entity.data.pipeline import (
     TaskStatus,
 )
 from metadata.generated.schema.entity.data.table import Table
+from metadata.generated.schema.entity.data.topic import Topic
 from metadata.generated.schema.entity.services.connections.pipeline.databricksPipelineConnection import (
     DatabricksPipelineConnection,
 )
@@ -56,6 +57,10 @@ from metadata.ingestion.api.steps import InvalidSourceException
 from metadata.ingestion.lineage.sql_lineage import get_column_fqn
 from metadata.ingestion.models.pipeline_status import OMetaPipelineStatus
 from metadata.ingestion.ometa.ometa_api import OpenMetadata
+from metadata.ingestion.source.pipeline.databrickspipeline.kafka_parser import (
+    extract_dlt_table_names,
+    extract_kafka_sources,
+)
 from metadata.ingestion.source.pipeline.databrickspipeline.models import (
     DataBrickPipelineDetails,
     DBRun,
@@ -86,6 +91,12 @@ class DatabrickspipelineSource(PipelineServiceSource):
     Pipeline metadata from Databricks Jobs API
     """
 
+    def __init__(self, config, metadata):
+        super().__init__(config, metadata)
+        # Cache for Databricks services to avoid repeated API calls
+        self._databricks_services_cached = False
+        self._databricks_services: List[str] = []
+
     @classmethod
     def create(
         cls, config_dict, metadata: OpenMetadata, pipeline_name: Optional[str] = None
@@ -105,14 +116,30 @@ class DatabrickspipelineSource(PipelineServiceSource):
                 yield DataBrickPipelineDetails(**workflow)
         except Exception as exc:
             logger.debug(traceback.format_exc())
-            logger.error(f"Failed to get pipeline list due to : {exc}")
+            logger.error(f"Failed to get jobs list due to : {exc}")
+
+        # Fetch DLT pipelines directly (new)
+        try:
+            for pipeline in self.client.list_pipelines() or []:
+                try:
+                    yield DataBrickPipelineDetails(**pipeline)
+                except Exception as exc:
+                    logger.debug(f"Error creating DLT pipeline details: {exc}")
+                    logger.debug(traceback.format_exc())
+                    continue
+        except Exception as exc:
+            logger.debug(traceback.format_exc())
+            logger.warning(f"Failed to get DLT pipelines list due to : {exc}")
+
         return None
 
     def get_pipeline_name(
         self, pipeline_details: DataBrickPipelineDetails
     ) -> Optional[str]:
         try:
-            return pipeline_details.settings.name
+            if pipeline_details.pipeline_id:
+                return pipeline_details.name
+            return pipeline_details.settings.name if pipeline_details.settings else None
         except Exception as exc:
             logger.debug(traceback.format_exc())
             logger.error(f"Failed to get pipeline name due to : {exc}")
@@ -124,17 +151,35 @@ class DatabrickspipelineSource(PipelineServiceSource):
     ) -> Iterable[Either[CreatePipelineRequest]]:
         """Method to Get Pipeline Entity"""
         try:
-            description = pipeline_details.settings.description
+            if pipeline_details.pipeline_id:
+                description = None
+                display_name = pipeline_details.name
+                entity_name = str(pipeline_details.pipeline_id)
+                schedule_interval = None
+            else:
+                description = (
+                    pipeline_details.settings.description
+                    if pipeline_details.settings
+                    else None
+                )
+                display_name = (
+                    pipeline_details.settings.name
+                    if pipeline_details.settings
+                    else None
+                )
+                entity_name = str(pipeline_details.job_id)
+                schedule_interval = (
+                    str(pipeline_details.settings.schedule.cron)
+                    if pipeline_details.settings and pipeline_details.settings.schedule
+                    else None
+                )
+
             pipeline_request = CreatePipelineRequest(
-                name=EntityName(str(pipeline_details.job_id)),
-                displayName=pipeline_details.settings.name,
+                name=EntityName(entity_name),
+                displayName=display_name,
                 description=Markdown(description) if description else None,
                 tasks=self.get_tasks(pipeline_details),
-                scheduleInterval=(
-                    str(pipeline_details.settings.schedule.cron)
-                    if pipeline_details.settings.schedule
-                    else None
-                ),
+                scheduleInterval=schedule_interval,
                 service=FullyQualifiedEntityName(self.context.get().pipeline_service),
             )
             yield Either(right=pipeline_request)
@@ -170,31 +215,25 @@ class DatabrickspipelineSource(PipelineServiceSource):
 
     def get_tasks(self, pipeline_details: DataBrickPipelineDetails) -> List[Task]:
         try:
-            task_list = []
-            for run in self.client.get_job_runs(job_id=pipeline_details.job_id) or []:
-                run = DBRun(**run)
-                task_list.extend(
-                    [
-                        Task(
-                            name=str(task.name),
-                            taskType=pipeline_details.settings.task_type,
-                            sourceUrl=(
-                                SourceUrl(run.run_page_url)
-                                if run.run_page_url
-                                else None
-                            ),
-                            description=(
-                                Markdown(task.description) if task.description else None
-                            ),
-                            downstreamTasks=[
-                                depend_task.name
-                                for depend_task in task.depends_on or []
-                            ],
-                        )
-                        for task in run.tasks or []
-                    ]
+            if not pipeline_details.settings or not pipeline_details.settings.tasks:
+                return None
+
+            job_url = f"https://{self.service_connection.hostPort}/#job/{pipeline_details.job_id}"
+
+            return [
+                Task(
+                    name=str(task.name),
+                    taskType=pipeline_details.settings.task_type,
+                    sourceUrl=SourceUrl(job_url),
+                    description=(
+                        Markdown(task.description) if task.description else None
+                    ),
+                    downstreamTasks=[
+                        depend_task.name for depend_task in task.depends_on or []
+                    ],
                 )
-            return task_list
+                for task in pipeline_details.settings.tasks
+            ]
         except Exception as exc:
             logger.debug(traceback.format_exc())
             logger.warning(f"Failed to get tasks list due to : {exc}")
@@ -204,6 +243,9 @@ class DatabrickspipelineSource(PipelineServiceSource):
         self, pipeline_details: DataBrickPipelineDetails
     ) -> Iterable[OMetaPipelineStatus]:
         try:
+            if not pipeline_details.job_id:
+                return
+
             for run in self.client.get_job_runs(job_id=pipeline_details.job_id) or []:
                 run = DBRun(**run)
                 task_status = [
@@ -241,7 +283,7 @@ class DatabrickspipelineSource(PipelineServiceSource):
         except Exception as exc:
             yield Either(
                 left=StackTraceError(
-                    name=pipeline_details.job_id,
+                    name=pipeline_details.id,
                     error=f"Failed to yield pipeline status: {exc}",
                     stackTrace=traceback.format_exc(),
                 )
@@ -300,6 +342,641 @@ class DatabrickspipelineSource(PipelineServiceSource):
             )
         return processed_column_lineage or []
 
+    def _get_databricks_services(self) -> List[str]:
+        """
+        Get list of all Databricks/Unity Catalog database service names from OpenMetadata
+
+        Caches the result to avoid repeated API calls during lineage extraction.
+        Returns list of service names that are of type Databricks or UnityCatalog.
+        """
+        # Return cached services if already fetched
+        if self._databricks_services_cached:
+            logger.debug(
+                f"Using cached Databricks services: {self._databricks_services}"
+            )
+            return self._databricks_services
+
+        try:
+            from metadata.generated.schema.entity.services.databaseService import (
+                DatabaseService,
+            )
+
+            logger.info("Fetching Databricks/Unity Catalog database services...")
+
+            # List all database services
+            services = self.metadata.list_all_entities(
+                entity=DatabaseService, fields=["serviceType"]
+            )
+
+            databricks_services = []
+            for service in services or []:
+                try:
+                    service_type = (
+                        service.serviceType.value
+                        if hasattr(service, "serviceType")
+                        else None
+                    )
+                    service_name = (
+                        service.name.root
+                        if hasattr(service.name, "root")
+                        else service.name
+                    )
+
+                    logger.debug(f"  Service: {service_name}, Type: {service_type}")
+
+                    # Check if it's a Databricks or Unity Catalog service
+                    if service_type and service_type.lower() in [
+                        "databricks",
+                        "unitycatalog",
+                    ]:
+                        databricks_services.append(service_name)
+                        logger.debug(
+                            f"    ✓ Databricks/Unity Catalog service: {service_name}"
+                        )
+
+                except Exception as exc:
+                    logger.debug(f"  Error processing service: {exc}")
+                    continue
+
+            # Cache the results
+            self._databricks_services = databricks_services
+            self._databricks_services_cached = True
+
+            logger.info(
+                f"Found {len(databricks_services)} Databricks/Unity Catalog service(s): {databricks_services}"
+            )
+            return databricks_services
+
+        except Exception as exc:
+            logger.warning(f"Error fetching Databricks services: {exc}")
+            logger.debug(traceback.format_exc())
+            # Cache empty list to avoid repeated failures
+            self._databricks_services = []
+            self._databricks_services_cached = True
+            return []
+
+    def _find_dlt_table(
+        self, table_name: str, catalog: Optional[str], schema: Optional[str]
+    ) -> Optional[Table]:
+        """
+        Find DLT table in OpenMetadata by iterating through Databricks services
+
+        DLT pipelines only write to Databricks/Unity Catalog Delta tables.
+        Uses catalog.schema.table_name from DLT spec to build FQN for each Databricks service.
+
+        Args:
+            table_name: Table name extracted from notebook code
+            catalog: Catalog name from DLT pipeline spec (database in OpenMetadata)
+            schema: Schema name from DLT pipeline spec
+
+        Returns:
+            Table entity if found, None otherwise
+        """
+        try:
+            logger.debug(
+                f"Searching for DLT table: catalog={catalog}, schema={schema}, table={table_name}"
+            )
+
+            # Get all Databricks/Unity Catalog services (uses cache)
+            databricks_services = self._get_databricks_services()
+
+            if not databricks_services:
+                logger.warning(
+                    "No Databricks/Unity Catalog services found in OpenMetadata"
+                )
+                # Fall back to configured dbServiceNames if available
+                databricks_services = self.get_db_service_names() or []
+                if databricks_services:
+                    logger.info(
+                        f"Using configured database services: {databricks_services}"
+                    )
+
+            if not databricks_services:
+                return None
+
+            logger.debug(f"Trying {len(databricks_services)} Databricks service(s)")
+
+            # Try each Databricks service with exact case
+            for service_name in databricks_services:
+                try:
+                    # Build FQN: service.catalog.schema.table
+                    table_fqn = fqn.build(
+                        metadata=self.metadata,
+                        entity_type=Table,
+                        service_name=service_name,
+                        database_name=catalog,
+                        schema_name=schema,
+                        table_name=table_name,
+                    )
+
+                    logger.debug(f"  Trying FQN: {table_fqn}")
+
+                    # Try to get table
+                    table = self.metadata.get_by_name(entity=Table, fqn=table_fqn)
+                    if table:
+                        logger.info(f"Found DLT table with FQN: {table_fqn}")
+                        return table
+
+                except Exception as exc:
+                    logger.debug(f"  Error checking service {service_name}: {exc}")
+                    continue
+
+            # If no exact match found, try case-insensitive (Unity Catalog lowercases table names)
+            logger.debug("Exact match not found, trying lowercase variants...")
+            for service_name in databricks_services:
+                try:
+                    table_fqn = fqn.build(
+                        metadata=self.metadata,
+                        entity_type=Table,
+                        service_name=service_name,
+                        database_name=catalog.lower() if catalog else None,
+                        schema_name=schema.lower() if schema else None,
+                        table_name=table_name.lower(),
+                    )
+
+                    logger.debug(f"  Trying lowercase FQN: {table_fqn}")
+
+                    table = self.metadata.get_by_name(entity=Table, fqn=table_fqn)
+                    if table:
+                        logger.info(
+                            f"Found DLT table with FQN (lowercase): {table_fqn}"
+                        )
+                        return table
+
+                except Exception as exc:
+                    logger.debug(
+                        f"  Error checking service {service_name} (lowercase): {exc}"
+                    )
+                    continue
+
+        except Exception as exc:
+            logger.debug(f"Could not find DLT table {table_name}: {exc}")
+            logger.debug(traceback.format_exc())
+
+        logger.warning(
+            f"DLT table not found: {catalog}.{schema}.{table_name}. "
+            f"Ensure the table is ingested from a Databricks/Unity Catalog database service."
+        )
+        return None
+
+    def _find_kafka_topic(self, topic_name: str) -> Optional[Topic]:
+        """
+        Find Kafka topic in OpenMetadata using Elasticsearch search
+
+        Handles topic names with dots (e.g., "dev.ern.cashout.moneyRequest_v1")
+        by searching with wildcard pattern: *.topic_name
+
+        Topic FQN format: MessagingServiceName.TopicName
+        When TopicName has dots, it's quoted: MessagingServiceName."dev.ern.topic"
+        """
+        try:
+            logger.debug(
+                f"Searching for topic {topic_name} across all messaging services"
+            )
+
+            # Use ES search with wildcard pattern to find topic regardless of service
+            # Pattern: *.topic_name or *."topic.with.dots"
+            from metadata.utils.elasticsearch import ES_INDEX_MAP
+
+            # Quote the topic name if it contains dots
+            search_topic_name = f'"{topic_name}"' if "." in topic_name else topic_name
+            search_pattern = f"*.{search_topic_name}"
+
+            logger.debug(f"Using search pattern: {search_pattern}")
+
+            # Search using ES field query for FQN pattern matching
+            query_string = f"/search/fieldQuery?fieldName=fullyQualifiedName&fieldValue={search_pattern}&from=0&size=10&index={ES_INDEX_MAP['Topic']}&deleted=false"
+
+            try:
+                response = self.metadata.client.get(query_string)
+                if response and response.get("hits", {}).get("hits"):
+                    # Get the first matching topic
+                    hit = response["hits"]["hits"][0]
+                    topic_fqn = hit["_source"]["fullyQualifiedName"]
+
+                    # Fetch full topic entity
+                    topic = self.metadata.get_by_name(entity=Topic, fqn=topic_fqn)
+                    if topic:
+                        logger.info(f"Found topic {topic_name} with FQN: {topic_fqn}")
+                        return topic
+            except Exception as search_exc:
+                logger.debug(f"ES search error: {search_exc}")
+                logger.debug(traceback.format_exc())
+
+        except Exception as exc:
+            logger.debug(f"Could not find topic {topic_name}: {exc}")
+            logger.debug(traceback.format_exc())
+
+        logger.warning(
+            f"Topic {topic_name} not found in OpenMetadata. "
+            f"Ensure the topic is ingested from a messaging service."
+        )
+        return None
+
+    def _yield_kafka_lineage(
+        self, pipeline_details: DataBrickPipelineDetails, pipeline_entity: Pipeline
+    ) -> Iterable[Either[AddLineageRequest]]:
+        """
+        Extract and yield Kafka topic lineage from DLT pipeline source code
+        Only processes DLT pipelines (with pipeline_id), not regular jobs
+        Creates lineage: Kafka topic -> DLT table (with pipeline in lineageDetails)
+        """
+        try:
+            logger.info("=" * 80)
+            logger.info(f"KAFKA LINEAGE EXTRACTION STARTED")
+            logger.info(
+                f"Pipeline: {pipeline_details.name if hasattr(pipeline_details, 'name') else 'N/A'}"
+            )
+            logger.info(f"Job ID: {pipeline_details.job_id}")
+            logger.info(f"Pipeline ID: {pipeline_details.pipeline_id}")
+            logger.info("=" * 80)
+
+            # Only process DLT pipelines - check for pipeline_id
+            # For pure DLT pipelines, pipeline_id is set directly
+            pipeline_id = pipeline_details.pipeline_id
+
+            # For jobs with DLT pipeline tasks, check settings
+            if not pipeline_id and pipeline_details.settings:
+                try:
+                    tasks = pipeline_details.settings.tasks
+                    logger.debug(
+                        f"Checking for DLT pipeline in job {pipeline_details.job_id}: "
+                        f"{len(tasks) if tasks else 0} tasks found"
+                    )
+
+                    if tasks:
+                        for task in tasks:
+                            logger.debug(
+                                f"Task: {task.name}, has pipeline_task: {task.pipeline_task is not None}"
+                            )
+                            # Check for direct DLT pipeline task
+                            if task.pipeline_task and task.pipeline_task.pipeline_id:
+                                pipeline_id = task.pipeline_task.pipeline_id
+                                logger.info(
+                                    f"✓ Found DLT pipeline_id from job task: {pipeline_id} for job {pipeline_details.job_id}"
+                                )
+                                break
+                except Exception as exc:
+                    logger.debug(f"Error checking for pipeline tasks: {exc}")
+                    logger.debug(traceback.format_exc())
+
+            # Only process if we have a DLT pipeline_id
+            if not pipeline_id:
+                logger.info(
+                    f"⊗ No DLT pipeline_id found - skipping Kafka lineage extraction"
+                )
+                logger.info(f"   Job ID: {pipeline_details.job_id}")
+                logger.info(f"   Pipeline ID: {pipeline_details.pipeline_id}")
+                logger.info("=" * 80)
+                return
+
+            logger.info(f"✓ Processing Kafka lineage for DLT pipeline: {pipeline_id}")
+
+            # Get pipeline configuration and extract target catalog/schema
+            target_catalog = None
+            target_schema = None
+            notebook_paths = []
+            try:
+                logger.info(f"⟳ Fetching pipeline configuration for {pipeline_id}...")
+                pipeline_config = self.client.get_pipeline_details(pipeline_id)
+                if not pipeline_config:
+                    logger.warning(
+                        f"✗ Could not fetch pipeline config for {pipeline_id}"
+                    )
+                    logger.info("=" * 80)
+                    return
+
+                logger.debug(f"✓ Pipeline config fetched successfully")
+                logger.debug(f"   Config keys: {list(pipeline_config.keys())}")
+
+                # Extract spec for detailed configuration
+                spec = pipeline_config.get("spec", {})
+                logger.info(f"✓ Pipeline spec extracted")
+                logger.info(f"   Spec keys: {list(spec.keys()) if spec else 'None'}")
+
+                # Extract target catalog and schema for DLT tables
+                target_catalog = spec.get("catalog") if spec else None
+                # Schema can be in 'target' or 'schema' field
+                target_schema = (
+                    spec.get("target") or spec.get("schema") if spec else None
+                )
+                logger.info(f"✓ DLT Target Location:")
+                logger.info(f"   Catalog: {target_catalog or 'NOT SET'}")
+                logger.info(f"   Schema: {target_schema or 'NOT SET'}")
+
+                # Extract notebook/file paths from libraries in spec
+                notebook_paths = []
+                if spec and "libraries" in spec:
+                    libraries = spec["libraries"]
+                    logger.info(
+                        f"⟳ Extracting notebook paths from {len(libraries)} libraries..."
+                    )
+                    for idx, lib in enumerate(libraries):
+                        logger.debug(f"   Library {idx + 1}: {lib}")
+                        # Library can be dict or have different structures
+                        if isinstance(lib, dict):
+                            # Check for notebook path
+                            if "notebook" in lib and lib["notebook"]:
+                                notebook = lib["notebook"]
+                                if isinstance(notebook, dict):
+                                    path = notebook.get("path")
+                                else:
+                                    path = notebook
+                                if path:
+                                    notebook_paths.append(path)
+                                    logger.info(f"   ✓ Found notebook: {path}")
+                            # Check for glob pattern
+                            elif "glob" in lib and lib["glob"]:
+                                glob_pattern = lib["glob"]
+                                if isinstance(glob_pattern, dict):
+                                    include_pattern = glob_pattern.get("include")
+                                    if include_pattern:
+                                        # Convert glob pattern to directory path
+                                        # e.g., "/path/**" -> "/path/"
+                                        base_path = include_pattern.replace(
+                                            "/**", "/"
+                                        ).replace("**", "")
+                                        notebook_paths.append(base_path)
+                                        logger.info(
+                                            f"   ✓ Found glob pattern, using base path: {base_path}"
+                                        )
+
+                # Also check for source path in spec configuration
+                if not notebook_paths and spec:
+                    source_path = None
+
+                    # Check spec.configuration for source path
+                    if "configuration" in spec:
+                        config = spec["configuration"]
+                        source_path = config.get("source_path") or config.get("source")
+
+                    # Check development settings
+                    if not source_path and "development" in spec:
+                        source_path = spec["development"].get("source_path")
+
+                    if source_path:
+                        logger.info(
+                            f"   ✓ Found source_path in pipeline spec: {source_path}"
+                        )
+                        notebook_paths.append(source_path)
+
+                logger.info(f"✓ Total notebook paths found: {len(notebook_paths)}")
+                for idx, path in enumerate(notebook_paths):
+                    logger.info(f"   {idx + 1}. {path}")
+            except Exception as exc:
+                logger.error(
+                    f"✗ Failed to fetch pipeline config for {pipeline_id}: {exc}"
+                )
+                logger.debug(traceback.format_exc())
+                logger.info("=" * 80)
+                return
+
+            if not notebook_paths:
+                logger.warning(f"✗ No notebook paths found for pipeline {pipeline_id}")
+                logger.info(
+                    "   Cannot extract Kafka lineage without notebook source code"
+                )
+                logger.info("=" * 80)
+                return
+
+            # Expand directories to individual notebook files
+            logger.info(f"⟳ Expanding directory paths to individual notebooks...")
+            expanded_paths = []
+            for path in notebook_paths:
+                # If path ends with /, it's a directory - list all notebooks in it
+                if path.endswith("/"):
+                    try:
+                        logger.debug(f"   Listing directory: {path}")
+                        # List workspace directory to get all notebooks
+                        objects = self.client.list_workspace_objects(path)
+                        if objects:
+                            for obj in objects:
+                                obj_type = obj.get("object_type")
+                                if obj_type in ("NOTEBOOK", "FILE"):
+                                    notebook_path = obj.get("path")
+                                    if notebook_path:
+                                        expanded_paths.append(notebook_path)
+                                        logger.info(
+                                            f"   ✓ Found {obj_type.lower()}: {notebook_path}"
+                                        )
+                        if not expanded_paths:
+                            logger.debug(f"   ⊗ No notebooks found in directory {path}")
+                    except Exception as exc:
+                        logger.warning(f"   ✗ Could not list directory {path}: {exc}")
+                        logger.debug(traceback.format_exc())
+                else:
+                    expanded_paths.append(path)
+                    logger.debug(f"   Direct path: {path}")
+
+            logger.info(f"✓ Total notebooks to process: {len(expanded_paths)}")
+
+            # Process each notebook to extract Kafka sources and DLT tables
+            logger.info("-" * 80)
+            logger.info(f"PROCESSING NOTEBOOKS FOR KAFKA LINEAGE")
+            logger.info("-" * 80)
+
+            for idx, lib_path in enumerate(expanded_paths, 1):
+                try:
+                    logger.info(f"\n📓 Notebook {idx}/{len(expanded_paths)}: {lib_path}")
+                    logger.info(f"⟳ Exporting notebook source code...")
+
+                    source_code = self.client.export_notebook_source(lib_path)
+                    if not source_code:
+                        logger.warning(f"✗ Could not export source for {lib_path}")
+                        continue
+
+                    logger.info(
+                        f"✓ Source code exported ({len(source_code)} characters)"
+                    )
+
+                    # Log full source code for debugging
+                    logger.debug(f"   ===== FULL NOTEBOOK SOURCE CODE =====")
+                    for i, line in enumerate(source_code.split("\n"), 1):
+                        logger.debug(f"   {i:3d}: {line}")
+                    logger.debug(f"   ===== END OF SOURCE CODE =====")
+
+                    # Extract Kafka topics
+                    logger.info(f"⟳ Parsing Kafka sources from notebook...")
+                    logger.debug(f"   Looking for patterns:")
+                    logger.debug(
+                        f"   - Kafka: .format('kafka')...option('subscribe', 'topic')"
+                    )
+                    logger.debug(f"   - DLT: @dlt.table(name='table_name')")
+                    kafka_sources = extract_kafka_sources(source_code)
+                    if kafka_sources:
+                        topics_found = [t for ks in kafka_sources for t in ks.topics]
+                        logger.info(
+                            f"✓ Found {len(kafka_sources)} Kafka source(s) with {len(topics_found)} topic(s):"
+                        )
+                        for ks_idx, ks in enumerate(kafka_sources, 1):
+                            logger.info(f"   Kafka Source {ks_idx}:")
+                            logger.info(f"     Topics: {ks.topics}")
+                            logger.info(
+                                f"     Bootstrap Servers: {ks.bootstrap_servers or 'NOT SET'}"
+                            )
+                            logger.info(
+                                f"     Group ID Prefix: {ks.group_id_prefix or 'NOT SET'}"
+                            )
+                    else:
+                        logger.info(f"⊗ No Kafka sources found in notebook")
+
+                    # Extract DLT table names
+                    logger.info(f"⟳ Parsing DLT table names from notebook...")
+                    dlt_table_names = extract_dlt_table_names(source_code)
+                    if dlt_table_names:
+                        logger.info(
+                            f"✓ Found {len(dlt_table_names)} DLT table(s): {dlt_table_names}"
+                        )
+                    else:
+                        logger.info(f"⊗ No DLT tables found in notebook")
+
+                    if not dlt_table_names or not kafka_sources:
+                        logger.warning(
+                            f"⊗ Skipping lineage for this notebook - need both Kafka sources AND DLT tables"
+                        )
+                        logger.info(
+                            f"   Kafka sources: {len(kafka_sources) if kafka_sources else 0}"
+                        )
+                        logger.info(
+                            f"   DLT tables: {len(dlt_table_names) if dlt_table_names else 0}"
+                        )
+                        continue
+
+                    logger.info(
+                        f"✓ Notebook has both Kafka sources and DLT tables - creating lineage..."
+                    )
+
+                    # Create lineage for each Kafka topic -> DLT table
+                    logger.info(f"\n⟳ Creating lineage edges...")
+                    lineage_created = 0
+
+                    for kafka_config in kafka_sources:
+                        for topic_name in kafka_config.topics:
+                            try:
+                                logger.info(f"\n   🔍 Processing topic: {topic_name}")
+
+                                # Use smart discovery to find topic
+                                logger.info(
+                                    f"   ⟳ Searching for topic in OpenMetadata..."
+                                )
+                                kafka_topic = self._find_kafka_topic(topic_name)
+
+                                if not kafka_topic:
+                                    logger.warning(
+                                        f"   ✗ Kafka topic '{topic_name}' not found in OpenMetadata"
+                                    )
+                                    logger.info(
+                                        f"   💡 Make sure the topic is ingested from a messaging service"
+                                    )
+                                    continue
+
+                                logger.info(
+                                    f"   ✓ Topic found: {kafka_topic.fullyQualifiedName.root if hasattr(kafka_topic.fullyQualifiedName, 'root') else kafka_topic.fullyQualifiedName}"
+                                )
+
+                                # Create lineage to each DLT table in this notebook
+                                for table_name in dlt_table_names:
+                                    logger.info(
+                                        f"   🔍 Processing target table: {table_name}"
+                                    )
+                                    logger.info(
+                                        f"   ⟳ Searching in Databricks/Unity Catalog services..."
+                                    )
+
+                                    # Use cached Databricks service lookup
+                                    target_table_entity = self._find_dlt_table(
+                                        table_name=table_name,
+                                        catalog=target_catalog,
+                                        schema=target_schema,
+                                    )
+
+                                    if target_table_entity:
+                                        table_fqn = (
+                                            target_table_entity.fullyQualifiedName.root
+                                            if hasattr(
+                                                target_table_entity.fullyQualifiedName,
+                                                "root",
+                                            )
+                                            else target_table_entity.fullyQualifiedName
+                                        )
+                                        logger.info(
+                                            f"   ✓ Target table found: {table_fqn}"
+                                        )
+                                        logger.info(
+                                            f"   ✅ Creating lineage: {topic_name} -> {table_fqn}"
+                                        )
+                                        logger.info(f"      Pipeline: {pipeline_id}")
+
+                                        yield Either(
+                                            right=AddLineageRequest(
+                                                edge=EntitiesEdge(
+                                                    fromEntity=EntityReference(
+                                                        id=kafka_topic.id,
+                                                        type="topic",
+                                                    ),
+                                                    toEntity=EntityReference(
+                                                        id=target_table_entity.id.root
+                                                        if hasattr(
+                                                            target_table_entity.id,
+                                                            "root",
+                                                        )
+                                                        else target_table_entity.id,
+                                                        type="table",
+                                                    ),
+                                                    lineageDetails=LineageDetails(
+                                                        pipeline=EntityReference(
+                                                            id=pipeline_entity.id.root,
+                                                            type="pipeline",
+                                                        ),
+                                                        source=LineageSource.PipelineLineage,
+                                                    ),
+                                                )
+                                            )
+                                        )
+                                        lineage_created += 1
+                                    else:
+                                        logger.warning(
+                                            f"   ✗ Target table '{table_name}' not found in OpenMetadata"
+                                        )
+                                        logger.info(
+                                            f"   💡 Expected location: {target_catalog}.{target_schema}.{table_name}"
+                                        )
+                                        logger.info(
+                                            f"   💡 Make sure the table is ingested from a Databricks/Unity Catalog database service"
+                                        )
+
+                            except Exception as exc:
+                                logger.error(
+                                    f"   ✗ Failed to process topic {topic_name}: {exc}"
+                                )
+                                logger.debug(traceback.format_exc())
+                                continue
+
+                    logger.info(
+                        f"\n✓ Lineage edges created for this notebook: {lineage_created}"
+                    )
+                except Exception as exc:
+                    logger.error(f"✗ Failed to process notebook {lib_path}: {exc}")
+                    logger.debug(traceback.format_exc())
+                    logger.info(f"   Continuing with next notebook...")
+                    continue
+
+            logger.info("\n" + "=" * 80)
+            logger.info(f"KAFKA LINEAGE EXTRACTION COMPLETED")
+            logger.info(
+                f"Pipeline: {pipeline_details.name if hasattr(pipeline_details, 'name') else 'N/A'}"
+            )
+            logger.info("=" * 80)
+
+        except Exception as exc:
+            logger.error(f"✗ Unexpected error in Kafka lineage extraction: {exc}")
+            logger.error(f"   Job ID: {pipeline_details.job_id}")
+            logger.error(
+                f"   Pipeline ID: {pipeline_details.pipeline_id if hasattr(pipeline_details, 'pipeline_id') else 'N/A'}"
+            )
+            logger.debug(traceback.format_exc())
+            logger.info("=" * 80)
+
     def yield_pipeline_lineage_details(
         self, pipeline_details: DataBrickPipelineDetails
     ) -> Iterable[Either[AddLineageRequest]]:
@@ -315,6 +992,13 @@ class DatabrickspipelineSource(PipelineServiceSource):
                 entity=Pipeline, fqn=pipeline_fqn
             )
 
+            # Extract Kafka topic lineage from source code
+            # Works automatically - no configuration required!
+            yield from self._yield_kafka_lineage(pipeline_details, pipeline_entity)
+
+            if not pipeline_details.job_id:
+                return
+
             table_lineage_list = self.client.get_table_lineage(
                 job_id=pipeline_details.job_id
             )
@@ -409,7 +1093,7 @@ class DatabrickspipelineSource(PipelineServiceSource):
         except Exception as exc:
             yield Either(
                 left=StackTraceError(
-                    name=pipeline_details.job_id,
+                    name=pipeline_details.id,
                     error=f"Wild error ingesting pipeline lineage {pipeline_details} - {exc}",
                     stackTrace=traceback.format_exc(),
                 )