feat: Trace pipeline run input/output data (#7590)

* Trace pipeline run * Add reno note * Update tracing tests to check input_data and output_data * empty --------- Co-authored-by: anakin87 <stefanofiorucci@gmail.com> Co-authored-by: Silvano Cerza <3314350+silvanocerza@users.noreply.github.com>
deepset-ai · Apr 29, 2024 · 8cb3cec · 8cb3cec
1 parent 451fae8
commit 8cb3cec
Show file tree

Hide file tree

Showing 3 changed files with 13 additions and 3 deletions.
diff --git a/haystack/core/pipeline/pipeline.py b/haystack/core/pipeline/pipeline.py
@@ -796,17 +796,19 @@ def run(self, word: str):
 
         include_outputs_from = set() if include_outputs_from is None else include_outputs_from
 
+        # This is what we'll return at the end
+        final_outputs: Dict[Any, Any] = {}
+
         with tracing.tracer.trace(
             "haystack.pipeline.run",
             tags={
+                "haystack.pipeline.input_data": data,
+                "haystack.pipeline.output_data": final_outputs,
                 "haystack.pipeline.debug": debug,
                 "haystack.pipeline.metadata": self.metadata,
                 "haystack.pipeline.max_loops_allowed": self.max_loops_allowed,
             },
         ):
-            # This is what we'll return at the end
-            final_outputs: Dict[Any, Any] = {}
-
             # Cache for extra outputs, if enabled.
             extra_outputs: Dict[Any, Any] = {}
 

diff --git a/releasenotes/notes/improve-pipeline-run-tracing-1d80aa7810df4205.yaml b/releasenotes/notes/improve-pipeline-run-tracing-1d80aa7810df4205.yaml
@@ -0,0 +1,4 @@
+---
+enhancements:
+ - |
+   Improved pipeline run tracing to include pipeline input/output data.
diff --git a/test/core/pipeline/test_tracing.py b/test/core/pipeline/test_tracing.py
@@ -39,6 +39,8 @@ def test_with_enabled_tracing(self, pipeline: Pipeline, spying_tracer: SpyingTra
             SpyingSpan(
                 operation_name="haystack.pipeline.run",
                 tags={
+                    "haystack.pipeline.input_data": {"hello": {"word": "world"}},
+                    "haystack.pipeline.output_data": {"hello2": {"output": "Hello, Hello, world!!"}},
                     "haystack.pipeline.debug": False,
                     "haystack.pipeline.metadata": {},
                     "haystack.pipeline.max_loops_allowed": 100,
@@ -97,6 +99,8 @@ def test_with_enabled_content_tracing(
                     "haystack.pipeline.debug": False,
                     "haystack.pipeline.metadata": {},
                     "haystack.pipeline.max_loops_allowed": 100,
+                    "haystack.pipeline.input_data": {"hello": {"word": "world"}},
+                    "haystack.pipeline.output_data": {"hello2": {"output": "Hello, Hello, world!!"}},
                 },
                 trace_id=ANY,
                 span_id=ANY,