refactor!: Remove of datadog tracer

MIGRATION.md

@@ -108,6 +108,54 @@ pip install <new-package>
 | `from haystack.components.embedders.image import SentenceTransformersDocumentImageEmbedder` | `sentence-transformers-haystack` | `from haystack_integrations.components.embedders.sentence_transformers import SentenceTransformersDocumentImageEmbedder` |
 | `from haystack.components.rankers import SentenceTransformersSimilarityRanker` | `sentence-transformers-haystack` | `from haystack_integrations.components.rankers.sentence_transformers import SentenceTransformersSimilarityRanker` |
 | `from haystack.components.rankers import SentenceTransformersDiversityRanker` | `sentence-transformers-haystack` | `from haystack_integrations.components.rankers.sentence_transformers import SentenceTransformersDiversityRanker` |
+| `from haystack.tracing.datadog import DatadogTracer` | `datadog-haystack` | `from haystack_integrations.tracing.datadog import DatadogTracer` |
+
+### `DatadogTracer` moved to the `datadog-haystack` integration
+
+**What changed:** The `DatadogTracer` has been moved out of Haystack into the `datadog-haystack` integration package.
+In addition, Haystack no longer automatically enables Datadog tracing when `ddtrace` is installed. You now enable it
+explicitly by adding the new `DatadogConnector` component to your pipeline.
+
+**Why:** Moving the tracer to a dedicated package keeps Haystack's dependencies leaner and lets the integration be
+released independently. Removing the implicit auto-enable makes tracing setup explicit and predictable.
+
+**How to migrate:**
+
+Install the integration:
+
+```bash
+pip install datadog-haystack
+```
+
+Before (v2.x), Datadog tracing was auto-enabled when `ddtrace` was installed, or set up manually:
+
+```python
+import ddtrace
+from haystack import tracing
+from haystack.tracing.datadog import DatadogTracer
+
+tracing.enable_tracing(DatadogTracer(ddtrace.tracer))
+```
+
+After (v3.0), add the `DatadogConnector` to your pipeline to enable tracing:
+
+```python
+from haystack import Pipeline
+from haystack_integrations.components.connectors.datadog import DatadogConnector
+
+pipe = Pipeline()
+pipe.add_component("tracer", DatadogConnector())
+```
+
+Alternatively, you can still enable the tracer manually using the new import path:
+
+```python
+import ddtrace
+from haystack import tracing
+from haystack_integrations.tracing.datadog import DatadogTracer
+
+tracing.enable_tracing(DatadogTracer(ddtrace.tracer))
+```
 
 ### `TransformersSimilarityRanker` removed
 

docs-website/docs/development/tracing.mdx

@@ -15,7 +15,7 @@ Traces document the flow of requests through your application and are vital for
 
 ## Configuring a Tracing Backend
 
-Instrumented applications typically send traces to a trace collector or a tracing backend. Haystack provides out-of-the-box support for [OpenTelemetry](https://opentelemetry.io/) and [Datadog](https://app.datadoghq.eu/dashboard/lists). You can also quickly implement support for additional providers of your choosing.
+Instrumented applications typically send traces to a trace collector or a tracing backend. Haystack provides out-of-the-box support for [OpenTelemetry](https://opentelemetry.io/) and, through integrations, for backends such as [Datadog](https://www.datadoghq.com/), [Langfuse](https://langfuse.com/), and [MLflow](https://mlflow.org/). You can also quickly implement support for additional providers of your choosing.
 
 ### OpenTelemetry
 
@@ -202,7 +202,6 @@ To use your custom tracing backend with Haystack, follow these steps:
 Haystack automatically detects and enables tracing under the following circumstances:
 
 - If `opentelemetry-sdk` is installed and configured for OpenTelemetry.
-- If `ddtrace` is installed for Datadog.
 
 To disable this behavior, there are two options:
 

haystack/tracing/tracer.py

@@ -198,7 +198,7 @@ def auto_enable_tracing() -> None:
     if is_tracing_enabled():
         return  # tracing already enabled
 
-    tracer = _auto_configured_opentelemetry_tracer() or _auto_configured_datadog_tracer()
+    tracer = _auto_configured_opentelemetry_tracer()
     if tracer:
         enable_tracing(tracer)
         logger.info("Auto-enabled tracing for '{tracer}'", tracer=tracer.__class__.__name__)
@@ -226,19 +226,4 @@ def _auto_configured_opentelemetry_tracer() -> Tracer | None:
     return None
 
 
-def _auto_configured_datadog_tracer() -> Tracer | None:
-    # we implement this here and not in the `datadog` module to avoid import warnings when Datadog is not installed
-    try:
-        from ddtrace.trace import tracer
-
-        from haystack.tracing.datadog import DatadogTracer
-
-        if tracer.enabled:
-            return DatadogTracer(tracer=tracer)
-    except ImportError:
-        pass
-
-    return None
-
-
 auto_enable_tracing()

pyproject.toml

@@ -129,7 +129,6 @@ dependencies = [
 
   # Tracing
   "opentelemetry-sdk",
-  "ddtrace",
 
   # Structured logging
   "structlog",

releasenotes/notes/remove-datadog-tracer-4f1c8a2b9d6e0a73.yaml

@@ -0,0 +1,36 @@
+---
+upgrade:
+  - |
+    The ``DatadogTracer`` has been moved out of Haystack into the ``datadog-haystack`` integration package.
+    It is no longer auto-enabled when ``ddtrace`` is installed; instead, use the new ``DatadogConnector`` component
+    to enable Datadog tracing in your pipeline. Install the new package with ``pip install datadog-haystack``.
+
+    Before:
+
+    .. code:: python
+
+      import ddtrace
+      from haystack import tracing
+      from haystack.tracing.datadog import DatadogTracer
+
+      tracing.enable_tracing(DatadogTracer(ddtrace.tracer))
+
+    After:
+
+    .. code:: python
+
+      from haystack import Pipeline
+      from haystack_integrations.components.connectors.datadog import DatadogConnector
+
+      pipe = Pipeline()
+      pipe.add_component("tracer", DatadogConnector())
+
+    Alternatively, you can still enable the tracer manually:
+
+    .. code:: python
+
+      import ddtrace
+      from haystack import tracing
+      from haystack_integrations.tracing.datadog import DatadogTracer
+
+      tracing.enable_tracing(DatadogTracer(ddtrace.tracer))