Wiring dimensional metrics

Author: TimPansino

newrelic/api/application.py

@@ -142,6 +142,14 @@ def record_custom_metrics(self, metrics):
         if self.active and metrics:
             self._agent.record_custom_metrics(self._name, metrics)
 
+    def record_dimensional_metric(self, name, value, tags=None):
+        if self.active:
+            self._agent.record_dimensional_metric(self._name, name, value, tags)
+
+    def record_dimensional_metrics(self, metrics):
+        if self.active and metrics:
+            self._agent.record_dimensional_metrics(self._name, metrics)
+
     def record_custom_event(self, event_type, params):
         if self.active:
             self._agent.record_custom_event(self._name, event_type, params)

newrelic/api/transaction.py

@@ -64,7 +64,7 @@
 from newrelic.core.custom_event import create_custom_event
 from newrelic.core.log_event_node import LogEventNode
 from newrelic.core.stack_trace import exception_stack
-from newrelic.core.stats_engine import CustomMetrics, SampledDataSet
+from newrelic.core.stats_engine import CustomMetrics, DimensionalMetrics, SampledDataSet
 from newrelic.core.thread_utilization import utilization_tracker
 from newrelic.core.trace_cache import (
     TraceCacheActiveTraceError,
@@ -307,6 +307,7 @@ def __init__(self, application, enabled=None, source=None):
         self.synthetics_header = None
 
         self._custom_metrics = CustomMetrics()
+        self._dimensional_metrics = DimensionalMetrics()
 
         global_settings = application.global_settings
 
@@ -588,6 +589,7 @@ def __exit__(self, exc, value, tb):
             apdex_t=self.apdex,
             suppress_apdex=self.suppress_apdex,
             custom_metrics=self._custom_metrics,
+            dimensional_metrics=self._dimensional_metrics,
             guid=self.guid,
             cpu_time=self._cpu_user_time_value,
             suppress_transaction_trace=self.suppress_transaction_trace,
@@ -1600,6 +1602,16 @@ def record_custom_metrics(self, metrics):
         for name, value in metrics:
             self._custom_metrics.record_custom_metric(name, value)
 
+    def record_dimensional_metric(self, name, value, tags=None):
+        self._dimensional_metrics.record_dimensional_metric(name, value, tags)
+
+    def record_dimensional_metrics(self, metrics):
+        for metric in metrics:
+            name, value = metric[:2]
+            tags = metric[2] if len(metric) >= 3 else None
+
+            self._dimensional_metrics.record_dimensional_metric(name, value, tags)
+
     def record_custom_event(self, event_type, params):
         settings = self._settings
 
@@ -1898,6 +1910,44 @@ def record_custom_metrics(metrics, application=None):
         application.record_custom_metrics(metrics)
 
 
+def record_dimensional_metric(name, value, tags=None, application=None):
+    if application is None:
+        transaction = current_transaction()
+        if transaction:
+            transaction.record_dimensional_metric(name, value, tags)
+        else:
+            _logger.debug(
+                "record_dimensional_metric has been called but no "
+                "transaction was running. As a result, the following metric "
+                "has not been recorded. Name: %r Value: %r Tags: %r. To correct this "
+                "problem, supply an application object as a parameter to this "
+                "record_dimensional_metrics call.",
+                name,
+                value,
+                tags,
+            )
+    elif application.enabled:
+        application.record_dimensional_metric(name, value, tags)
+
+
+def record_dimensional_metrics(metrics, application=None):
+    if application is None:
+        transaction = current_transaction()
+        if transaction:
+            transaction.record_dimensional_metrics(metrics)
+        else:
+            _logger.debug(
+                "record_dimensional_metrics has been called but no "
+                "transaction was running. As a result, the following metrics "
+                "have not been recorded: %r. To correct this problem, "
+                "supply an application object as a parameter to this "
+                "record_dimensional_metric call.",
+                list(metrics),
+            )
+    elif application.enabled:
+        application.record_dimensional_metrics(metrics)
+
+
 def record_custom_event(event_type, params, application=None):
     """Record a custom event.
 

newrelic/common/metric_utils.py

@@ -0,0 +1,31 @@
+# Copyright 2010 New Relic, Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+This module implements functions for creating a unique identity from a name and set of tags for use in dimensional metrics.
+"""
+
+from newrelic.packages import six
+
+
+def create_metric_identity(name, tags=None):
+    if tags:
+        if isinstance(tags, dict):
+            tags = frozenset(six.iteritems(tags)) if tags is not None else None
+        elif not isinstance(tags, frozenset):
+            tags = frozenset(tags)
+    elif tags is not None:
+        tags = None  # Set empty iterables to None
+
+    return (name, tags)

newrelic/core/agent.py

@@ -524,6 +524,33 @@ def record_custom_metrics(self, app_name, metrics):
 
         application.record_custom_metrics(metrics)
 
+    def record_dimensional_metric(self, app_name, name, value, tags=None):
+        """Records a basic metric for the named application. If there has
+        been no prior request to activate the application, the metric is
+        discarded.
+
+        """
+
+        application = self._applications.get(app_name, None)
+        if application is None or not application.active:
+            return
+
+        application.record_dimensional_metric(name, value, tags)
+
+    def record_dimensional_metrics(self, app_name, metrics):
+        """Records the metrics for the named application. If there has
+        been no prior request to activate the application, the metric is
+        discarded. The metrics should be an iterable yielding tuples
+        consisting of the name and value.
+
+        """
+
+        application = self._applications.get(app_name, None)
+        if application is None or not application.active:
+            return
+
+        application.record_dimensional_metrics(metrics)
+
     def record_custom_event(self, app_name, event_type, params):
         application = self._applications.get(app_name, None)
         if application is None or not application.active:

newrelic/core/application.py

@@ -106,6 +106,9 @@ def __init__(self, app_name, linked_applications=None):
         self._stats_custom_lock = threading.RLock()
         self._stats_custom_engine = StatsEngine()
 
+        self._stats_dimensional_lock = threading.RLock()
+        self._stats_dimensional_engine = StatsEngine()
+
         self._agent_commands_lock = threading.Lock()
         self._data_samplers_lock = threading.Lock()
         self._data_samplers_started = False
@@ -510,6 +513,9 @@ def connect_to_data_collector(self, activate_agent):
         with self._stats_custom_lock:
             self._stats_custom_engine.reset_stats(configuration)
 
+        with self._stats_dimensional_lock:
+            self._stats_dimensional_engine.reset_stats(configuration)
+
         # Record an initial start time for the reporting period and
         # clear record of last transaction processed.
 
@@ -860,6 +866,50 @@ def record_custom_metrics(self, metrics):
                 self._global_events_account += 1
                 self._stats_custom_engine.record_custom_metric(name, value)
 
+    def record_dimensional_metric(self, name, value, tags=None):
+        """Record a dimensional metric against the application independent
+        of a specific transaction.
+
+        NOTE that this will require locking of the stats engine for
+        dimensional metrics and so under heavy use will have performance
+        issues. It is better to record the dimensional metric against an
+        active transaction as they will then be aggregated at the end of
+        the transaction when all other metrics are aggregated and so no
+        additional locking will be required.
+
+        """
+
+        if not self._active_session:
+            return
+
+        with self._stats_dimensional_lock:
+            self._global_events_account += 1
+            self._stats_dimensional_engine.record_dimensional_metric(name, value, tags)
+
+    def record_dimensional_metrics(self, metrics):
+        """Record a set of dimensional metrics against the application
+        independent of a specific transaction.
+
+        NOTE that this will require locking of the stats engine for
+        dimensional metrics and so under heavy use will have performance
+        issues. It is better to record the dimensional metric against an
+        active transaction as they will then be aggregated at the end of
+        the transaction when all other metrics are aggregated and so no
+        additional locking will be required.
+
+        """
+
+        if not self._active_session:
+            return
+
+        with self._stats_dimensional_lock:
+            for metric in metrics:
+                name, value = metric[:2]
+                tags = metric[2] if len(metric) >= 3 else None
+
+                self._global_events_account += 1
+                self._stats_dimensional_engine.record_dimensional_metric(name, value, tags)
+
     def record_custom_event(self, event_type, params):
         if not self._active_session:
             return
@@ -1416,11 +1466,14 @@ def harvest(self, shutdown=False, flexible=False):
                         _logger.debug("Normalizing metrics for harvest of %r.", self._app_name)
 
                         metric_data = stats.metric_data(metric_normalizer)
+                        dimensional_metric_data = stats.dimensional_metric_data(metric_normalizer)
 
                         _logger.debug("Sending metric data for harvest of %r.", self._app_name)
 
                         # Send metrics
                         self._active_session.send_metric_data(self._period_start, period_end, metric_data)
+                        if dimensional_metric_data:
+                            self._active_session.send_dimensional_metric_data(self._period_start, period_end, dimensional_metric_data)
 
                         _logger.debug("Done sending data for harvest of %r.", self._app_name)
 

newrelic/core/data_collector.py

@@ -31,6 +31,8 @@
 
 _logger = logging.getLogger(__name__)
 
+DIMENSIONAL_METRIC_DATA_TEMP = []  # TODO: REMOVE THIS
+
 
 class Session(object):
     PROTOCOL = AgentProtocol
@@ -128,6 +130,25 @@ def send_metric_data(self, start_time, end_time, metric_data):
         payload = (self.agent_run_id, start_time, end_time, metric_data)
         return self._protocol.send("metric_data", payload)
 
+    def send_dimensional_metric_data(self, start_time, end_time, metric_data):
+        """Called to submit dimensional metric data for specified period of time.
+        Time values are seconds since UNIX epoch as returned by the
+        time.time() function. The metric data should be iterable of
+        specific metrics.
+
+        NOTE: This data is sent not sent to the normal agent endpoints but is sent
+        to the MELT API endpoints to keep the entity separate. This is for use
+        with the machine learning integration only.
+        """
+
+        payload = (self.agent_run_id, start_time, end_time, metric_data)
+        # return self._protocol.send("metric_data", payload)
+
+        # TODO: REMOVE THIS. Replace with actual protocol.
+        DIMENSIONAL_METRIC_DATA_TEMP.append(payload)
+        _logger.debug("Dimensional Metrics: %r" % metric_data)
+        return 200  
+
     def send_log_events(self, sampling_info, log_event_data):
         """Called to submit sample set for log events."""