Merge pull request #76 from autometrics-dev/add-service-name

Add service.name

Author: actualwitch

Diff for: CHANGELOG.md

@@ -22,6 +22,10 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 - When the `function.calls.duration` histogram is exported to Prometheus, it now
   includes the units (`function_calls_duration_seconds`) to be in line with
   Prometheus/OpenMetrics naming conventions. **Dashboards and alerting rules must be updated.**
+- The `caller` label on the `function.calls` metric was replaced with `caller.function`
+  and `caller.module`
+- All metrics now have a `service.name` label attached. This is set via runtime environment
+  variable (`AUTOMETRICS_SERVICE_NAME` or `OTEL_SERVICE_NAME`), or falls back to the package name.
 
 ### Deprecated
 

Diff for: README.md

@@ -89,7 +89,7 @@ def api_handler():
 
 Autometrics keeps track of instrumented functions calling each other. If you have a function that calls another function, metrics for later will include `caller` label set to the name of the autometricised function that called it.
 
-#### Metrics Libraries
+## Metrics Libraries
 
 Configure the package that autometrics will use to produce metrics with the `AUTOMETRICS_TRACKER` environment variable.
 
@@ -116,6 +116,17 @@ Configure the labels by setting the following environment variables:
 
 This follows the method outlined in [Exposing the software version to Prometheus](https://www.robustperception.io/exposing-the-software-version-to-prometheus/).
 
+## Service name
+
+All metrics produced by Autometrics have a label called `service.name` (or `service_name` when exported to Prometheus) attached to identify the logical service they are part of.
+
+You may want to override the default service name, for example if you are running multiple instances of the same code base as separate services and want to differentiate between the metrics produced by each one.
+
+The service name is loaded from the following environment variables, in this order:
+
+1. `AUTOMETRICS_SERVICE_NAME` (at runtime)
+2. `OTEL_SERVICE_NAME` (at runtime)
+
 ## Exemplars
 
 > **NOTE** - As of writing, exemplars aren't supported by the default tracker (`AUTOMETRICS_TRACKER=OPEN_TELEMETRY`).

Diff for: src/autometrics/constants.py

@@ -5,11 +5,13 @@
 CONCURRENCY_NAME = "function.calls.concurrent"
 # NOTE - The Rust implementation does not use `build.info`, instead opts for just `build_info`
 BUILD_INFO_NAME = "build_info"
+SERVICE_NAME = "service.name"
 
 
 COUNTER_NAME_PROMETHEUS = COUNTER_NAME.replace(".", "_")
 HISTOGRAM_NAME_PROMETHEUS = HISTOGRAM_NAME.replace(".", "_")
 CONCURRENCY_NAME_PROMETHEUS = CONCURRENCY_NAME.replace(".", "_")
+SERVICE_NAME_PROMETHEUS = SERVICE_NAME.replace(".", "_")
 
 COUNTER_DESCRIPTION = "Autometrics counter for tracking function calls"
 HISTOGRAM_DESCRIPTION = "Autometrics histogram for tracking function call duration"

Diff for: src/autometrics/test_caller.py

@@ -38,6 +38,6 @@ def bar():
     assert blob is not None
     data = blob.decode("utf-8")
 
-    expected = """function_calls_total{caller_function="test_caller_detection.<locals>.bar",caller_module="autometrics.test_caller",function="test_caller_detection.<locals>.foo",module="autometrics.test_caller",objective_name="",objective_percentile="",result="ok"} 1.0"""
+    expected = """function_calls_total{caller_function="test_caller_detection.<locals>.bar",caller_module="autometrics.test_caller",function="test_caller_detection.<locals>.foo",module="autometrics.test_caller",objective_name="",objective_percentile="",result="ok",service_name="autometrics"} 1.0"""
     assert "wrapper" not in data
     assert expected in data

Diff for: src/autometrics/test_decorator.py

@@ -24,12 +24,12 @@
     HISTOGRAM_NAME,
     BUILD_INFO_NAME,
     BUILD_INFO_DESCRIPTION,
+    SERVICE_NAME,
     OBJECTIVE_NAME,
     OBJECTIVE_PERCENTILE,
     OBJECTIVE_LATENCY_THRESHOLD,
 )
-
-FUNCTION_CALLS_DURATION_NAME = "function.calls.duration"
+from ..utils import get_service_name
 
 
 def get_objective_boundaries():
@@ -103,6 +103,7 @@ def __count(
                 "caller.function": caller_function,
                 OBJECTIVE_NAME: objective_name,
                 OBJECTIVE_PERCENTILE: percentile,
+                SERVICE_NAME: get_service_name(),
             },
         )
 
@@ -130,6 +131,7 @@ def __histogram(
             attributes={
                 "function": function,
                 "module": module,
+                SERVICE_NAME: get_service_name(),
                 OBJECTIVE_NAME: objective_name,
                 OBJECTIVE_PERCENTILE: percentile,
                 OBJECTIVE_LATENCY_THRESHOLD: threshold,
@@ -145,11 +147,15 @@ def set_build_info(self, commit: str, version: str, branch: str):
                     "commit": commit,
                     "version": version,
                     "branch": branch,
+                    SERVICE_NAME: get_service_name(),
                 },
             )
 
     def start(
-        self, function: str, module: str, track_concurrency: Optional[bool] = False
+        self,
+        function: str,
+        module: str,
+        track_concurrency: Optional[bool] = False,
     ):
         """Start tracking metrics for a function call."""
         if track_concurrency:
@@ -158,6 +164,7 @@ def start(
                 attributes={
                     "function": function,
                     "module": module,
+                    SERVICE_NAME: get_service_name(),
                 },
             )
 
@@ -195,6 +202,7 @@ def finish(
                 attributes={
                     "function": function,
                     "module": module,
+                    SERVICE_NAME: get_service_name(),
                 },
             )
 

Diff for: src/autometrics/tracker/opentelemetry.py

@@ -7,6 +7,7 @@
     COUNTER_NAME_PROMETHEUS,
     HISTOGRAM_NAME_PROMETHEUS,
     CONCURRENCY_NAME_PROMETHEUS,
+    SERVICE_NAME_PROMETHEUS,
     BUILD_INFO_NAME,
     COUNTER_DESCRIPTION,
     HISTOGRAM_DESCRIPTION,
@@ -23,6 +24,7 @@
 from .exemplar import get_exemplar
 from .tracker import Result
 from ..objectives import Objective
+from ..utils import get_service_name
 
 
 class PrometheusTracker:
@@ -34,6 +36,7 @@ class PrometheusTracker:
         [
             "function",
             "module",
+            SERVICE_NAME_PROMETHEUS,
             "result",
             "caller_module",
             "caller_function",
@@ -47,17 +50,26 @@ class PrometheusTracker:
         [
             "function",
             "module",
+            SERVICE_NAME_PROMETHEUS,
             OBJECTIVE_NAME_PROMETHEUS,
             OBJECTIVE_PERCENTILE_PROMETHEUS,
             OBJECTIVE_LATENCY_THRESHOLD_PROMETHEUS,
         ],
         unit="seconds",
     )
     prom_gauge_build_info = Gauge(
-        BUILD_INFO_NAME, BUILD_INFO_DESCRIPTION, [COMMIT_KEY, VERSION_KEY, BRANCH_KEY]
+        BUILD_INFO_NAME,
+        BUILD_INFO_DESCRIPTION,
+        [COMMIT_KEY, VERSION_KEY, BRANCH_KEY, SERVICE_NAME_PROMETHEUS],
     )
     prom_gauge_concurrency = Gauge(
-        CONCURRENCY_NAME_PROMETHEUS, CONCURRENCY_DESCRIPTION, ["function", "module"]
+        CONCURRENCY_NAME_PROMETHEUS,
+        CONCURRENCY_DESCRIPTION,
+        [
+            "function",
+            "module",
+            SERVICE_NAME_PROMETHEUS,
+        ],
     )
 
     def __init__(self) -> None:
@@ -81,10 +93,12 @@ def _count(
             if objective is None or objective.success_rate is None
             else objective.success_rate.value
         )
+        service_name = get_service_name()
 
         self.prom_counter.labels(
             func_name,
             module_name,
+            service_name,
             result.value,
             caller_module,
             caller_function,
@@ -110,10 +124,12 @@ def _histogram(
         if latency is not None:
             threshold = latency[0].value
             percentile = latency[1].value
+        service_name = get_service_name()
 
         self.prom_histogram.labels(
             func_name,
             module_name,
+            service_name,
             objective_name,
             percentile,
             threshold,
@@ -122,14 +138,18 @@ def _histogram(
     def set_build_info(self, commit: str, version: str, branch: str):
         if not self._has_set_build_info:
             self._has_set_build_info = True
-            self.prom_gauge_build_info.labels(commit, version, branch).set(1)
+            service_name = get_service_name()
+            self.prom_gauge_build_info.labels(
+                commit, version, branch, service_name
+            ).set(1)
 
     def start(
         self, function: str, module: str, track_concurrency: Optional[bool] = False
     ):
         """Start tracking metrics for a function call."""
         if track_concurrency:
-            self.prom_gauge_concurrency.labels(function, module).inc()
+            service_name = get_service_name()
+            self.prom_gauge_concurrency.labels(function, module, service_name).inc()
 
     def finish(
         self,
@@ -159,7 +179,8 @@ def finish(
         self._histogram(function, module, start_time, objective, exemplar)
 
         if track_concurrency:
-            self.prom_gauge_concurrency.labels(function, module).dec()
+            service_name = get_service_name()
+            self.prom_gauge_concurrency.labels(function, module, service_name).dec()
 
     def initialize_counters(
         self,

Diff for: src/autometrics/tracker/prometheus.py

@@ -37,7 +37,7 @@ async def test_concurrency_tracking_prometheus(monkeypatch):
     data = blob.decode("utf-8")
 
     assert (
-        f"""# TYPE function_calls_concurrent gauge\nfunction_calls_concurrent{{function="sleep",module="autometrics.tracker.test_concurrency"}} 1.0"""
+        f"""# TYPE function_calls_concurrent gauge\nfunction_calls_concurrent{{function="sleep",module="autometrics.tracker.test_concurrency",service_name="autometrics"}} 1.0"""
         in data
     )
 

Diff for: src/autometrics/tracker/test_concurrency.py

@@ -44,10 +44,9 @@ def test_init_prometheus_tracker_set_build_info(monkeypatch):
     blob = generate_latest()
     assert blob is not None
     data = blob.decode("utf-8")
+    print(data)
 
-    prom_build_info = (
-        f"""build_info{{branch="{branch}",commit="{commit}",version="{version}"}} 1.0"""
-    )
+    prom_build_info = f"""build_info{{branch="{branch}",commit="{commit}",service_name="autometrics",version="{version}"}} 1.0"""
     assert prom_build_info in data
 
     monkeypatch.delenv("AUTOMETRICS_VERSION", raising=False)
@@ -79,9 +78,7 @@ def test_init_otel_tracker_set_build_info(monkeypatch):
     assert blob is not None
     data = blob.decode("utf-8")
 
-    prom_build_info = (
-        f"""build_info{{branch="{branch}",commit="{commit}",version="{version}"}} 1.0"""
-    )
+    prom_build_info = f"""build_info{{branch="{branch}",commit="{commit}",service_name="autometrics",version="{version}",service_name="autometrics"}} 1.0"""
     assert prom_build_info in data
 
     monkeypatch.delenv("AUTOMETRICS_VERSION", raising=False)

Diff for: src/autometrics/tracker/test_tracker.py

@@ -29,6 +29,21 @@ def get_function_name(func: Callable) -> str:
     return func.__qualname__ or func.__name__
 
 
+service_name = None
+
+
+def get_service_name():
+    global service_name
+    if service_name is not None:
+        return service_name
+    service_name = (
+        os.getenv("AUTOMETRICS_SERVICE_NAME")
+        or os.getenv("OTEL_SERVICE_NAME")
+        or __package__.rsplit(".", 1)[0]
+    )
+    return service_name
+
+
 def write_docs(func_name: str, module_name: str):
     """Write the prometheus query urls to the function docstring."""
     generator = Generator(func_name, module_name)