metrics: Remove LabelSet (#527)

Remove LabelSet according to OTEP 90:
https://github.com/open-telemetry/oteps/blob/master/text/0090-remove-labelset-from-metrics-api.md

Co-Authored-By: Leighton Chen <[email protected]>

docs/examples/basic_meter/basic_metrics.py

@@ -84,21 +84,21 @@ def usage(argv):
     label_keys=("environment",),
 )
 
-# Labelsets are used to identify key-values that are associated with a specific
+# Labels are used to identify key-values that are associated with a specific
 # metric that you want to record. These are useful for pre-aggregation and can
 # be used to store custom dimensions pertaining to a metric
-staging_label_set = meter.get_label_set({"environment": "staging"})
-testing_label_set = meter.get_label_set({"environment": "testing"})
+staging_labels = {"environment": "staging"}
+testing_labels = {"environment": "testing"}
 
 # Update the metric instruments using the direct calling convention
-requests_counter.add(25, staging_label_set)
-requests_size.record(100, staging_label_set)
+requests_counter.add(25, staging_labels)
+requests_size.record(100, staging_labels)
 time.sleep(5)
 
-requests_counter.add(50, staging_label_set)
-requests_size.record(5000, staging_label_set)
+requests_counter.add(50, staging_labels)
+requests_size.record(5000, staging_labels)
 time.sleep(5)
 
-requests_counter.add(35, testing_label_set)
-requests_size.record(2, testing_label_set)
+requests_counter.add(35, testing_labels)
+requests_size.record(2, testing_labels)
 time.sleep(5)

docs/examples/basic_meter/calling_conventions.py

@@ -56,27 +56,27 @@
     label_keys=("environment",),
 )
 
-label_set = meter.get_label_set({"environment": "staging"})
+labels = {"environment": "staging"}
 
 print("Updating using direct calling convention...")
-# You can record metrics directly using the metric instrument. You pass in a
-# labelset that you would like to record for.
-requests_counter.add(25, label_set)
+# You can record metrics directly using the metric instrument. You pass in
+# labels that you would like to record for.
+requests_counter.add(25, labels)
 time.sleep(5)
 
 print("Updating using a bound instrument...")
 # You can record metrics with bound metric instruments. Bound metric
-# instruments are created by passing in a labelset. A bound metric instrument
+# instruments are created by passing in labels. A bound metric instrument
 # is essentially metric data that corresponds to a specific set of labels.
 # Therefore, getting a bound metric instrument using the same set of labels
 # will yield the same bound metric instrument.
-bound_requests_counter = requests_counter.bind(label_set)
+bound_requests_counter = requests_counter.bind(labels)
 bound_requests_counter.add(100)
 time.sleep(5)
 
 print("Updating using batch calling convention...")
-# You can record metrics in a batch by passing in a labelset and a sequence of
+# You can record metrics in a batch by passing in labels and a sequence of
 # (metric, value) pairs. The value would be recorded for each metric using the
-# specified labelset for each.
-meter.record_batch(label_set, ((requests_counter, 50), (clicks_counter, 70)))
+# specified labels for each.
+meter.record_batch(labels, ((requests_counter, 50), (clicks_counter, 70)))
 time.sleep(5)

docs/examples/basic_meter/observer.py

@@ -19,7 +19,7 @@
 import psutil
 
 from opentelemetry import metrics
-from opentelemetry.sdk.metrics import LabelSet, MeterProvider
+from opentelemetry.sdk.metrics import MeterProvider
 from opentelemetry.sdk.metrics.export import ConsoleMetricsExporter
 from opentelemetry.sdk.metrics.export.batcher import UngroupedBatcher
 from opentelemetry.sdk.metrics.export.controller import PushController
@@ -35,8 +35,8 @@
 # Callback to gather cpu usage
 def get_cpu_usage_callback(observer):
     for (number, percent) in enumerate(psutil.cpu_percent(percpu=True)):
-        label_set = meter.get_label_set({"cpu_number": str(number)})
-        observer.observe(percent, label_set)
+        labels = {"cpu_number": str(number)}
+        observer.observe(percent, labels)
 
 
 meter.register_observer(
@@ -52,7 +52,7 @@ def get_cpu_usage_callback(observer):
 # Callback to gather RAM memory usage
 def get_ram_usage_callback(observer):
     ram_percent = psutil.virtual_memory().percent
-    observer.observe(ram_percent, LabelSet())
+    observer.observe(ram_percent, {})
 
 
 meter.register_observer(

docs/examples/otcollector-metrics/collector.py

@@ -40,8 +40,8 @@
     label_keys=("environment",),
 )
 
-staging_label_set = meter.get_label_set({"environment": "staging"})
-requests_counter.add(25, staging_label_set)
+staging_labels = {"environment": "staging"}
+requests_counter.add(25, staging_labels)
 
 print("Metrics are available now at http://localhost:9090/graph")
 input("Press any key to exit...")

docs/examples/prometheus/prometheus.py

@@ -41,8 +41,8 @@
     label_keys=("environment",),
 )
 
-staging_label_set = meter.get_label_set({"environment": "staging"})
-requests_counter.add(25, staging_label_set)
+staging_labels = {"environment": "staging"}
+requests_counter.add(25, staging_labels)
 
 print("Metrics are available now at http://localhost:8000/")
 input("Press any key to exit...")

docs/getting-started.rst

@@ -249,7 +249,7 @@ The following is an example of emitting metrics to console, in a similar fashion
     exporter = ConsoleMetricsExporter()
     controller = PushController(meter, exporter, 5)
 
-    staging_label_set = meter.get_label_set({"environment": "staging"})
+    staging_labels = {"environment": "staging"}
 
     requests_counter = meter.create_metric(
         name="requests",
@@ -260,10 +260,10 @@ The following is an example of emitting metrics to console, in a similar fashion
         label_keys=("environment",),
     )
 
-    requests_counter.add(25, staging_label_set)
+    requests_counter.add(25, staging_labels)
     time.sleep(5)
 
-    requests_counter.add(20, staging_label_set)
+    requests_counter.add(20, staging_labels)
     time.sleep(5)
 
 
@@ -272,8 +272,8 @@ The sleeps will cause the script to take a while, but running it should yield:
 .. code-block:: sh
 
     $ python metrics.py
-    ConsoleMetricsExporter(data="Counter(name="requests", description="number of requests")", label_set="(('environment', 'staging'),)", value=25)
-    ConsoleMetricsExporter(data="Counter(name="requests", description="number of requests")", label_set="(('environment', 'staging'),)", value=45)
+    ConsoleMetricsExporter(data="Counter(name="requests", description="number of requests")", labels="(('environment', 'staging'),)", value=25)
+    ConsoleMetricsExporter(data="Counter(name="requests", description="number of requests")", labels="(('environment', 'staging'),)", value=45)
 
 Using Prometheus
 ----------------
@@ -331,7 +331,7 @@ And use that instead of the `ConsoleMetricsExporter`:
     exporter = PrometheusMetricsExporter("MyAppPrefix")
     controller = PushController(meter, exporter, 5)
 
-    staging_label_set = meter.get_label_set({"environment": "staging"})
+    staging_labels = {"environment": "staging"}
 
     requests_counter = meter.create_metric(
         name="requests",
@@ -342,10 +342,10 @@ And use that instead of the `ConsoleMetricsExporter`:
         label_keys=("environment",),
     )
 
-    requests_counter.add(25, staging_label_set)
+    requests_counter.add(25, staging_labels)
     time.sleep(5)
 
-    requests_counter.add(20, staging_label_set)
+    requests_counter.add(20, staging_labels)
     time.sleep(5)
 
     # This line is added to keep the HTTP server up long enough to scrape.
@@ -463,11 +463,9 @@ And execute the following script:
         metric_type=Counter,
         label_keys=("environment",),
     )
-
-    # Labelsets are used to identify key-values that are associated with a specific
+    # Labels are used to identify key-values that are associated with a specific
     # metric that you want to record. These are useful for pre-aggregation and can
     # be used to store custom dimensions pertaining to a metric
-    label_set = meter.get_label_set({"environment": "staging"})
-
-    requests_counter.add(25, label_set)
+    labels = {"environment": "staging"}
+    requests_counter.add(25, labels)
     time.sleep(10)  # give push_controller time to push metrics

docs/metrics_example.py

@@ -17,7 +17,7 @@
     label_keys=("environment",),
 )
 
-staging_label_set = meter.get_label_set({"environment": "staging"})
-requests_counter.add(25, staging_label_set)
+staging_labels = {"environment": "staging"}
+requests_counter.add(25, staging_labels)
 
 input("Press a key to finish...\n")

ext/opentelemetry-ext-otcollector/src/opentelemetry/ext/otcollector/__init__.py

@@ -77,10 +77,10 @@
         Counter,
         ("environment",),
     )
-    # Labelsets are used to identify key-values that are associated with a specific
+    # Labels are used to identify key-values that are associated with a specific
     # metric that you want to record. These are useful for pre-aggregation and can
     # be used to store custom dimensions pertaining to a metric
-    label_set = meter.get_label_set({"environment": "staging"})
+    labels = {"environment": "staging"}
 
-    counter.add(25, label_set)
+    counter.add(25, labels)
 """

ext/opentelemetry-ext-otcollector/src/opentelemetry/ext/otcollector/metrics_exporter/__init__.py

@@ -106,7 +106,7 @@ def translate_to_collector(
 
         label_values = []
         label_keys = []
-        for label_tuple in metric_record.label_set.labels:
+        for label_tuple in metric_record.labels:
             label_keys.append(metrics_pb2.LabelKey(key=label_tuple[0]))
             label_values.append(
                 metrics_pb2.LabelValue(
@@ -145,11 +145,12 @@ def get_collector_metric_type(metric: Metric) -> metrics_pb2.MetricDescriptor:
 
 
 def get_collector_point(metric_record: MetricRecord) -> metrics_pb2.Point:
+    # TODO: horrible hack to get original list of keys to then get the bound
+    # instrument
+    key = dict(metric_record.labels)
     point = metrics_pb2.Point(
         timestamp=utils.proto_timestamp_from_time_ns(
-            metric_record.metric.bind(
-                metric_record.label_set
-            ).last_update_timestamp
+            metric_record.metric.bind(key).last_update_timestamp
         )
     )
     if metric_record.metric.value_type == int:

ext/opentelemetry-ext-otcollector/tests/test_otcollector_metrics_exporter.py

@@ -21,7 +21,12 @@
 
 from opentelemetry import metrics
 from opentelemetry.ext.otcollector import metrics_exporter
-from opentelemetry.sdk.metrics import Counter, Measure, MeterProvider
+from opentelemetry.sdk.metrics import (
+    Counter,
+    Measure,
+    MeterProvider,
+    get_labels_as_key,
+)
 from opentelemetry.sdk.metrics.export import (
     MetricRecord,
     MetricsExportResult,
@@ -36,8 +41,8 @@ def setUpClass(cls):
         # pylint: disable=protected-access
         metrics.set_meter_provider(MeterProvider())
         cls._meter = metrics.get_meter(__name__)
-        kvp = {"environment": "staging"}
-        cls._test_label_set = cls._meter.get_label_set(kvp)
+        cls._labels = {"environment": "staging"}
+        cls._key_labels = get_labels_as_key(cls._labels)
 
     def test_constructor(self):
         mock_get_node = mock.Mock()
@@ -77,7 +82,6 @@ def test_get_collector_metric_type(self):
 
     def test_get_collector_point(self):
         aggregator = aggregate.CounterAggregator()
-        label_set = self._meter.get_label_set({"environment": "staging"})
         int_counter = self._meter.create_metric(
             "testName", "testDescription", "unit", int, Counter
         )
@@ -88,21 +92,21 @@ def test_get_collector_point(self):
             "testName", "testDescription", "unit", float, Measure
         )
         result = metrics_exporter.get_collector_point(
-            MetricRecord(aggregator, label_set, int_counter)
+            MetricRecord(aggregator, self._key_labels, int_counter)
         )
         self.assertIsInstance(result, metrics_pb2.Point)
         self.assertIsInstance(result.timestamp, Timestamp)
         self.assertEqual(result.int64_value, 0)
         aggregator.update(123.5)
         aggregator.take_checkpoint()
         result = metrics_exporter.get_collector_point(
-            MetricRecord(aggregator, label_set, float_counter)
+            MetricRecord(aggregator, self._key_labels, float_counter)
         )
         self.assertEqual(result.double_value, 123.5)
         self.assertRaises(
             TypeError,
             metrics_exporter.get_collector_point(
-                MetricRecord(aggregator, label_set, measure)
+                MetricRecord(aggregator, self._key_labels, measure)
             ),
         )
 
@@ -118,7 +122,7 @@ def test_export(self):
             "testname", "testdesc", "unit", int, Counter, ["environment"]
         )
         record = MetricRecord(
-            aggregate.CounterAggregator(), self._test_label_set, test_metric
+            aggregate.CounterAggregator(), self._key_labels, test_metric
         )
 
         result = collector_exporter.export([record])
@@ -137,14 +141,13 @@ def test_export(self):
         )
 
     def test_translate_to_collector(self):
-
         test_metric = self._meter.create_metric(
             "testname", "testdesc", "unit", int, Counter, ["environment"]
         )
         aggregator = aggregate.CounterAggregator()
         aggregator.update(123)
         aggregator.take_checkpoint()
-        record = MetricRecord(aggregator, self._test_label_set, test_metric)
+        record = MetricRecord(aggregator, self._key_labels, test_metric)
         output_metrics = metrics_exporter.translate_to_collector([record])
         self.assertEqual(len(output_metrics), 1)
         self.assertIsInstance(output_metrics[0], metrics_pb2.Metric)
@@ -175,12 +178,12 @@ def test_translate_to_collector(self):
         self.assertEqual(len(output_metrics[0].timeseries[0].points), 1)
         self.assertEqual(
             output_metrics[0].timeseries[0].points[0].timestamp.seconds,
-            record.metric.bind(record.label_set).last_update_timestamp
+            record.metric.bind(self._labels).last_update_timestamp
             // 1000000000,
         )
         self.assertEqual(
             output_metrics[0].timeseries[0].points[0].timestamp.nanos,
-            record.metric.bind(record.label_set).last_update_timestamp
+            record.metric.bind(self._labels).last_update_timestamp
             % 1000000000,
         )
         self.assertEqual(

ext/opentelemetry-ext-prometheus/src/opentelemetry/ext/prometheus/__init__.py

@@ -54,12 +54,12 @@
         ("environment",),
     )
 
-    # Labelsets are used to identify key-values that are associated with a specific
+    # Labels are used to identify key-values that are associated with a specific
     # metric that you want to record. These are useful for pre-aggregation and can
     # be used to store custom dimensions pertaining to a metric
-    label_set = meter.get_label_set({"environment": "staging"})
+    labels = {"environment": "staging"}
 
-    counter.add(25, label_set)
+    counter.add(25, labels)
     input("Press any key to exit...")
 
 API
@@ -145,7 +145,7 @@ def _translate_to_prometheus(self, metric_record: MetricRecord):
         prometheus_metric = None
         label_values = []
         label_keys = []
-        for label_tuple in metric_record.label_set.labels:
+        for label_tuple in metric_record.labels:
             label_keys.append(self._sanitize(label_tuple[0]))
             label_values.append(label_tuple[1])