From e6f524f95179661dfddba5c2d35d47e7de89f504 Mon Sep 17 00:00:00 2001
From: Yuneng Jiang <yuneng@berri.ai>
Date: Mon, 4 May 2026 20:35:09 -0700
Subject: [PATCH] [Fix] Tests: Pick chat-completion OTEL trace by content, not
 recency

The /otel-spans endpoint returns process-wide spans and tags
most_recent_parent by max start_time. After tightening that route to
proxy_admin (sk-1234), the GET /otel-spans request itself emits auth
spans that beat the chat-completion spans on start_time, so
most_recent_parent now points at the request's own auth trace
(['postgres', 'postgres']) and the >=5-span assertion fails.

Pick the chat-completion trace by content: it is the only trace whose
span list is a superset of {postgres, redis, raw_gen_ai_request,
batch_write_to_db}. Verified locally end-to-end against
otel_test_config.yaml + OTEL_EXPORTER=in_memory: 3/3 runs green.
---
 tests/otel_tests/test_otel.py | 22 ++++++++++++++++++----
 1 file changed, 18 insertions(+), 4 deletions(-)

diff --git a/tests/otel_tests/test_otel.py b/tests/otel_tests/test_otel.py
index a9ed4461a1..a0f58dd5b8 100644
--- a/tests/otel_tests/test_otel.py
+++ b/tests/otel_tests/test_otel.py
@@ -104,10 +104,24 @@ async def test_chat_completion_check_otel_spans():
         print("otel_spans: ", otel_spans)
 
         all_otel_spans = otel_spans["otel_spans"]
-        most_recent_parent = str(otel_spans["most_recent_parent"])
-        print("Most recent OTEL parent: ", most_recent_parent)
-        print("\n spans grouped by parent: ", otel_spans["spans_grouped_by_parent"])
-        parent_trace_spans = otel_spans["spans_grouped_by_parent"][most_recent_parent]
+        spans_grouped_by_parent = otel_spans["spans_grouped_by_parent"]
+        print("\n spans grouped by parent: ", spans_grouped_by_parent)
+
+        # The GET /otel-spans request itself produces auth spans that beat
+        # the chat-completion spans on start_time, so `most_recent_parent`
+        # points at the wrong trace. Pick the chat-completion trace by
+        # content: it's the one carrying the full set of expected markers.
+        chat_completion_markers = {
+            "postgres",
+            "redis",
+            "raw_gen_ai_request",
+            "batch_write_to_db",
+        }
+        parent_trace_spans = next(
+            spans
+            for spans in spans_grouped_by_parent.values()
+            if chat_completion_markers.issubset(spans)
+        )
 
         print("Parent trace spans: ", parent_trace_spans)