Add auto detection for application ID from connection string if not set (#44644)

* Add Auto-Detection of Application ID to Resource Attributes

* Fix format

* Add CHANGELOG

* No need for value assigning statement

* Modularize fetching appid logic and fix format

* Fix format

* Randomize conn string

Author: rads-1996

sdk/monitor/azure-monitor-opentelemetry-exporter/CHANGELOG.md

@@ -3,6 +3,8 @@
 ## 1.0.0b47 (Unreleased)
 
 ### Features Added
+- Add auto detection for application ID from connection string if not set 
+  ([#44644](https://github.com/Azure/azure-sdk-for-python/pull/44644))
 - Add support for user id and authId
 ([#44662](https://github.com/Azure/azure-sdk-for-python/pull/44662))
 

sdk/monitor/azure-monitor-opentelemetry-exporter/azure/monitor/opentelemetry/exporter/_connection_string_parser.py

@@ -9,6 +9,7 @@
 INSTRUMENTATION_KEY = "instrumentationkey"
 # cspell:disable-next-line
 AAD_AUDIENCE = "aadaudience"
+APPLICATION_ID = "applicationid"  # cspell:disable-line
 
 # Validate UUID format
 # Specs taken from https://tools.ietf.org/html/rfc4122
@@ -38,6 +39,7 @@ def __init__(self, connection_string: typing.Optional[str] = None) -> None:
         self._connection_string = connection_string
         self.aad_audience = ""
         self.region = ""
+        self.application_id = ""
         self._initialize()
         self._validate_instrumentation_key()
 
@@ -73,6 +75,9 @@ def _initialize(self) -> None:
         # Extract region information
         self.region = self._extract_region()  # type: ignore
 
+        # Extract application_id
+        self.application_id = code_cs.get(APPLICATION_ID) or env_cs.get(APPLICATION_ID)  # type: ignore
+
     def _extract_region(self) -> typing.Optional[str]:
         """Extract region from endpoint URL.
 

sdk/monitor/azure-monitor-opentelemetry-exporter/azure/monitor/opentelemetry/exporter/_constants.py

@@ -336,4 +336,7 @@ class _RP_Names(Enum):
 # Default message for messages(MessageData) with empty body
 _DEFAULT_LOG_MESSAGE = "n/a"
 
+# Resource attribute applicationId
+_APPLICATION_ID_RESOURCE_KEY = "microsoft.applicationId"
+
 # cSpell:disable

sdk/monitor/azure-monitor-opentelemetry-exporter/azure/monitor/opentelemetry/exporter/_utils.py

@@ -23,6 +23,7 @@
     TelemetryItem,
 )
 from azure.monitor.opentelemetry.exporter._version import VERSION as ext_version
+from azure.monitor.opentelemetry.exporter._connection_string_parser import ConnectionStringParser
 from azure.monitor.opentelemetry.exporter._constants import (
     _AKS_ARM_NAMESPACE_ID,
     _DEFAULT_AAD_SCOPE,
@@ -438,3 +439,8 @@ def get_compute_type():
 
 def _get_sha256_hash(input_str: str) -> str:
     return hashlib.sha256(input_str.encode("utf-8")).hexdigest()
+
+
+def _get_application_id(connection_string: Optional[str]) -> Optional[str]:
+    parsed_connection_string = ConnectionStringParser(connection_string)
+    return parsed_connection_string.application_id

sdk/monitor/azure-monitor-opentelemetry-exporter/azure/monitor/opentelemetry/exporter/export/trace/_exporter.py

@@ -4,7 +4,7 @@
 import json
 import logging
 from time import time_ns
-from typing import no_type_check, Any, Dict, List, Sequence
+from typing import no_type_check, Any, Dict, List, Sequence, Optional
 from urllib.parse import urlparse
 
 from opentelemetry.semconv.attributes.client_attributes import CLIENT_ADDRESS
@@ -38,6 +38,7 @@
     _REQUEST_ENVELOPE_NAME,
     _EXCEPTION_ENVELOPE_NAME,
     _REMOTE_DEPENDENCY_ENVELOPE_NAME,
+    _APPLICATION_ID_RESOURCE_KEY,
 )
 from azure.monitor.opentelemetry.exporter import _utils
 from azure.monitor.opentelemetry.exporter._generated.models import (
@@ -124,6 +125,7 @@ class AzureMonitorTraceExporter(BaseExporter, SpanExporter):
     def __init__(self, **kwargs: Any):
         self._tracer_provider = kwargs.pop("tracer_provider", None)
         super().__init__(**kwargs)
+        self.application_id = _utils._get_application_id(self._connection_string)
 
     def export(self, spans: Sequence[ReadableSpan], **_kwargs: Any) -> SpanExportResult:
         """Export span data.
@@ -134,12 +136,13 @@ def export(self, spans: Sequence[ReadableSpan], **_kwargs: Any) -> SpanExportRes
         :rtype: ~opentelemetry.sdk.trace.export.SpanExportResult
         """
         envelopes = []
+
         if spans and self._should_collect_otel_resource_metric():
             resource = None
             try:
                 tracer_provider = self._tracer_provider or get_tracer_provider()
                 resource = tracer_provider.resource  # type: ignore
-                envelopes.append(self._get_otel_resource_envelope(resource))
+                envelopes.append(self._get_otel_resource_envelope(resource, self.application_id))
             except AttributeError as e:
                 _logger.exception("Failed to derive Resource from Tracer Provider: %s", e)  # pylint: disable=C4769
         for span in spans:
@@ -162,14 +165,18 @@ def shutdown(self) -> None:
             self.storage.close()
 
     # pylint: disable=protected-access
-    def _get_otel_resource_envelope(self, resource: Resource) -> TelemetryItem:
-        attributes: Dict[str, str] = {}
+    def _get_otel_resource_envelope(self, resource: Resource, application_id: Optional[str]) -> TelemetryItem:
+        attributes: Dict[str, Any] = {}
         if resource:
-            attributes = resource.attributes
+            attributes = dict(resource.attributes)
         envelope = _utils._create_telemetry_item(time_ns())
         envelope.name = _METRIC_ENVELOPE_NAME
         envelope.tags.update(_utils._populate_part_a_fields(resource))  # pylint: disable=W0212
         envelope.instrumentation_key = self._instrumentation_key
+
+        if application_id and attributes.get(_APPLICATION_ID_RESOURCE_KEY) is None:
+            attributes[_APPLICATION_ID_RESOURCE_KEY] = application_id
+
         data_point = MetricDataPoint(
             name="_OTELRESOURCE_"[:1024],
             value=0,

sdk/monitor/azure-monitor-opentelemetry-exporter/tests/logs/test_logs.py

@@ -32,6 +32,7 @@
     _APPLICATION_INSIGHTS_EVENT_MARKER_ATTRIBUTE,
     _MICROSOFT_CUSTOM_EVENT_NAME,
     _DEFAULT_LOG_MESSAGE,
+    _APPLICATION_ID_RESOURCE_KEY,
 )
 from azure.monitor.opentelemetry.exporter._generated.models import ContextTagKeys
 from azure.monitor.opentelemetry.exporter._utils import (
@@ -128,7 +129,7 @@ def setUpClass(cls):
                 body=None,
                 attributes={"test": "attribute"},
             ),
-            resource=Resource.create(attributes={"asd": "test_resource"}),
+            resource=Resource.create(attributes={"asd": "test_resource", "microsoft.applicationId": "app_id"}),
             instrumentation_scope=InstrumentationScope("test_name"),
         )
         cls._log_data_complex_body = _logs.ReadWriteLogRecord(
@@ -470,6 +471,7 @@ def test_log_to_envelope_log_none(self):
         self.assertEqual(envelope.name, "Microsoft.ApplicationInsights.Message")
         self.assertEqual(envelope.data.base_type, "MessageData")
         self.assertEqual(envelope.data.base_data.message, _DEFAULT_LOG_MESSAGE)
+        self.assertEqual(envelope.data.base_data.properties.get(_APPLICATION_ID_RESOURCE_KEY), None)
 
     def test_log_to_envelope_log_empty(self):
         exporter = self._exporter
@@ -783,7 +785,6 @@ def test_get_severity_level(self):
         for sev_num in SeverityNumber:
             num = sev_num.value
             level = _get_severity_level(sev_num)
-            print(num)
             if num in range(0, 9):
                 self.assertEqual(level, 0)
             elif num in range(9, 13):

sdk/monitor/azure-monitor-opentelemetry-exporter/tests/test_connection_string_parser.py

@@ -357,3 +357,21 @@ def test_region_extraction_alphanumeric_regions(self):
                     + f";IngestionEndpoint=https://{endpoint_suffix}"
                 )
                 self.assertEqual(parser.region, expected_region)
+
+    def test_application_id_extraction_from_connection_string(self):
+        parser = ConnectionStringParser(
+            connection_string="InstrumentationKey="
+            + self._valid_instrumentation_key
+            + ";IngestionEndpoint=https://northeurope-999.in.applicationinsights.azure.com/"
+            + ";LiveEndpoint=https://eastus.livediagnostics.monitor.azure.com/;ApplicationId=3cd3dd3f-64cc-4d7c-9303-8d69a4bb8558"
+        )
+        self.assertEqual(parser.application_id, "3cd3dd3f-64cc-4d7c-9303-8d69a4bb8558")
+
+    def test_application_id_extraction_from_no_application_id(self):
+        parser = ConnectionStringParser(
+            connection_string="InstrumentationKey="
+            + self._valid_instrumentation_key
+            + ";IngestionEndpoint=https://northeurope-999.in.applicationinsights.azure.com/"
+            + ";LiveEndpoint=https://eastus.livediagnostics.monitor.azure.com/"
+        )
+        self.assertEqual(parser.application_id, None)

sdk/monitor/azure-monitor-opentelemetry-exporter/tests/trace/test_trace.py

@@ -41,6 +41,7 @@
     _AZURE_SDK_NAMESPACE_NAME,
     _AZURE_SDK_OPENTELEMETRY_NAME,
     _AZURE_AI_SDK_NAME,
+    _APPLICATION_ID_RESOURCE_KEY,
 )
 from azure.monitor.opentelemetry.exporter._generated.models import ContextTagKeys
 from azure.monitor.opentelemetry.exporter._utils import azure_monitor_context
@@ -133,6 +134,7 @@ def test_export_failure(self):
             transmit.return_value = ExportResult.FAILED_RETRYABLE
             storage_mock = mock.Mock()
             exporter.storage.put = storage_mock
+            exporter._connection_string = "InstrumentationKey=4321abcd-5678-4efa-8abc-1234567890ab;IngestionEndpoint=https://eastus-8.in.applicationinsights.azure.com/;LiveEndpoint=https://eastus.livediagnostics.monitor.azure.com/;ApplicationId=4321abcd-5678-4efa-8abc-1234567890ab"
             result = exporter.export([test_span])
         self.assertEqual(result, SpanExportResult.FAILURE)
         self.assertEqual(storage_mock.call_count, 1)
@@ -188,7 +190,7 @@ def test_export_with_tracer_provider(self):
                 "azure.monitor.opentelemetry.exporter.AzureMonitorTraceExporter._get_otel_resource_envelope"
             ) as resource_patch:  # noqa: E501
                 result = exporter.export([test_span])
-                resource_patch.assert_called_once_with(mock_resource)
+                resource_patch.assert_called_once_with(mock_resource, None)
                 self.assertEqual(result, SpanExportResult.SUCCESS)
                 self.assertEqual(storage_mock.call_count, 1)
 
@@ -221,7 +223,7 @@ def test_export_with_tracer_provider_global(self):
                 "azure.monitor.opentelemetry.exporter.AzureMonitorTraceExporter._get_otel_resource_envelope"
             ) as resource_patch:  # noqa: E501
                 result = exporter.export([test_span])
-                resource_patch.assert_called_once_with(mock_resource)
+                resource_patch.assert_called_once_with(mock_resource, None)
                 self.assertEqual(result, SpanExportResult.SUCCESS)
                 self.assertEqual(storage_mock.call_count, 1)
 
@@ -1768,7 +1770,7 @@ def test_export_otel_resource_metric(self, mock_get_tracer_provider):
                 mock_get_otel_resource_envelope.return_value = "test_envelope"
                 result = exporter.export([test_span])
                 self.assertEqual(result, SpanExportResult.SUCCESS)
-                mock_get_otel_resource_envelope.assert_called_once_with(test_resource)
+                mock_get_otel_resource_envelope.assert_called_once_with(test_resource, None)
                 envelopes = ["test_envelope", exporter._span_to_envelope(test_span)]
                 transmit.assert_called_once_with(envelopes)
 
@@ -1781,9 +1783,10 @@ def test_get_otel_resource_envelope(self):
                 "bool_test_key": False,
                 "float_test_key": 0.5,
                 "sequence_test_key": ["a", "b"],
+                "microsoft.applicationId": "test_app_id",
             }
         )
-        envelope = exporter._get_otel_resource_envelope(test_resource)
+        envelope = exporter._get_otel_resource_envelope(test_resource, "test_app_id")
         metric_name = envelope.name
         self.assertEqual(metric_name, "Microsoft.ApplicationInsights.Metric")
         instrumentation_key = envelope.instrumentation_key