Skip to content

feat(telemetry): add cache usage metrics to OpenTelemetry spans #825

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Sep 11, 2025
Merged

feat(telemetry): add cache usage metrics to OpenTelemetry spans #825

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions src/strands/telemetry/tracer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -263,6 +263,8 @@ def end_model_invoke_span(
"gen_ai.usage.completion_tokens": usage["outputTokens"],
"gen_ai.usage.output_tokens": usage["outputTokens"],
"gen_ai.usage.total_tokens": usage["totalTokens"],
"gen_ai.usage.cache_read_input_tokens": usage.get("cacheReadInputTokens", 0),
"gen_ai.usage.cache_write_input_tokens": usage.get("cacheWriteInputTokens", 0),
}

self._add_event(
Expand Down Expand Up @@ -491,6 +493,8 @@ def end_agent_span(
"gen_ai.usage.input_tokens": accumulated_usage["inputTokens"],
"gen_ai.usage.output_tokens": accumulated_usage["outputTokens"],
"gen_ai.usage.total_tokens": accumulated_usage["totalTokens"],
"gen_ai.usage.cache_read_input_tokens": accumulated_usage.get("cacheReadInputTokens", 0),
"gen_ai.usage.cache_write_input_tokens": accumulated_usage.get("cacheWriteInputTokens", 0),
}
)

Expand Down
62 changes: 62 additions & 0 deletions tests/strands/telemetry/test_tracer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -177,6 +177,8 @@ def test_end_model_invoke_span(mock_span):
mock_span.set_attribute.assert_any_call("gen_ai.usage.completion_tokens", 20)
mock_span.set_attribute.assert_any_call("gen_ai.usage.output_tokens", 20)
mock_span.set_attribute.assert_any_call("gen_ai.usage.total_tokens", 30)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_read_input_tokens", 0)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_write_input_tokens", 0)
mock_span.add_event.assert_called_with(
"gen_ai.choice",
attributes={"message": json.dumps(message["content"]), "finish_reason": "end_turn"},
Expand Down Expand Up @@ -404,6 +406,8 @@ def test_end_agent_span(mock_span):
mock_span.set_attribute.assert_any_call("gen_ai.usage.completion_tokens", 100)
mock_span.set_attribute.assert_any_call("gen_ai.usage.output_tokens", 100)
mock_span.set_attribute.assert_any_call("gen_ai.usage.total_tokens", 150)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_read_input_tokens", 0)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_write_input_tokens", 0)
mock_span.add_event.assert_any_call(
"gen_ai.choice",
attributes={"message": "Agent response", "finish_reason": "end_turn"},
Expand All @@ -412,6 +416,64 @@ def test_end_agent_span(mock_span):
mock_span.end.assert_called_once()


def test_end_model_invoke_span_with_cache_metrics(mock_span):
"""Test ending a model invoke span with cache metrics."""
tracer = Tracer()
message = {"role": "assistant", "content": [{"text": "Response"}]}
usage = Usage(
inputTokens=10,
outputTokens=20,
totalTokens=30,
cacheReadInputTokens=5,
cacheWriteInputTokens=3,
)
stop_reason: StopReason = "end_turn"

tracer.end_model_invoke_span(mock_span, message, usage, stop_reason)

mock_span.set_attribute.assert_any_call("gen_ai.usage.prompt_tokens", 10)
mock_span.set_attribute.assert_any_call("gen_ai.usage.input_tokens", 10)
mock_span.set_attribute.assert_any_call("gen_ai.usage.completion_tokens", 20)
mock_span.set_attribute.assert_any_call("gen_ai.usage.output_tokens", 20)
mock_span.set_attribute.assert_any_call("gen_ai.usage.total_tokens", 30)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_read_input_tokens", 5)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_write_input_tokens", 3)
mock_span.set_status.assert_called_once_with(StatusCode.OK)
mock_span.end.assert_called_once()


def test_end_agent_span_with_cache_metrics(mock_span):
"""Test ending an agent span with cache metrics."""
tracer = Tracer()

# Mock AgentResult with metrics including cache tokens
mock_metrics = mock.MagicMock()
mock_metrics.accumulated_usage = {
"inputTokens": 50,
"outputTokens": 100,
"totalTokens": 150,
"cacheReadInputTokens": 25,
"cacheWriteInputTokens": 10,
}

mock_response = mock.MagicMock()
mock_response.metrics = mock_metrics
mock_response.stop_reason = "end_turn"
mock_response.__str__ = mock.MagicMock(return_value="Agent response")

tracer.end_agent_span(mock_span, mock_response)

mock_span.set_attribute.assert_any_call("gen_ai.usage.prompt_tokens", 50)
mock_span.set_attribute.assert_any_call("gen_ai.usage.input_tokens", 50)
mock_span.set_attribute.assert_any_call("gen_ai.usage.completion_tokens", 100)
mock_span.set_attribute.assert_any_call("gen_ai.usage.output_tokens", 100)
mock_span.set_attribute.assert_any_call("gen_ai.usage.total_tokens", 150)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_read_input_tokens", 25)
mock_span.set_attribute.assert_any_call("gen_ai.usage.cache_write_input_tokens", 10)
mock_span.set_status.assert_called_once_with(StatusCode.OK)
mock_span.end.assert_called_once()


def test_get_tracer_singleton():
"""Test that get_tracer returns a singleton instance."""
# Reset the singleton first
Expand Down
Loading