feat(detectors): Add filtered_paths setting for Large HTTP Detector (#100842)

follow up to #100634 - we want
to allow users to explicitly state any paths that we should filter out
if we see them in spans. gives the user a bit more control beyond any
automatic filtering we do. will need a frontend PR to allow users to
edit these settings and update the detection algorithm to support this
filtering

Author: roggenkemper

src/sentry/features/temporary.py

@@ -173,7 +173,7 @@ def register_temporary_features(manager: FeatureManager) -> None:
     manager.add("organizations:metric-issue-poc", OrganizationFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=True)
     manager.add("projects:metric-issue-creation", ProjectFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=False)
     # Enable Large HTTP Payload Detector Improvements
-    manager.add("organizations:large-http-payload-detector-improvements", OrganizationFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=False)
+    manager.add("organizations:large-http-payload-detector-improvements", OrganizationFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=True)
     manager.add("organizations:mep-rollout-flag", OrganizationFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=True)
     manager.add("organizations:mep-use-default-tags", OrganizationFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=False)
     manager.add("organizations:disable-clustering-setting", OrganizationFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=True, default=False)

src/sentry/issue_detection/performance_detection.py

@@ -176,6 +176,9 @@ def get_merged_settings(project_id: int | None = None) -> dict[str | Any, Any]:
         "large_http_payload_size_threshold": options.get(
             "performance.issues.large_http_payload.size_threshold"
         ),
+        "large_http_payload_filtered_paths": options.get(
+            "performance.issues.large_http_payload.filtered_paths"
+        ),
         "db_on_main_thread_duration_threshold": options.get(
             "performance.issues.db_on_main_thread.total_spans_duration_threshold"
         ),
@@ -329,6 +332,7 @@ def get_detection_settings(
             "detection_enabled": settings["large_http_payload_detection_enabled"],
             "minimum_span_duration": 100,  # ms
             "organization": organization,
+            "filtered_paths": settings["large_http_payload_filtered_paths"],
         },
         DetectorType.HTTP_OVERHEAD: {
             "http_request_delay_threshold": settings["http_request_delay_threshold"],

src/sentry/issues/endpoints/project_performance_issue_settings.py

@@ -57,6 +57,7 @@ class ConfigurableThresholds(Enum):
     UNCOMPRESSED_ASSET_SIZE = "uncompressed_asset_size_threshold"
     LARGE_HTTP_PAYLOAD = "large_http_payload_detection_enabled"
     LARGE_HTTP_PAYLOAD_SIZE = "large_http_payload_size_threshold"
+    LARGE_HTTP_PAYLOAD_FILTERED_PATHS = "large_http_payload_filtered_paths"
     DB_ON_MAIN_THREAD = "db_on_main_thread_detection_enabled"
     DB_ON_MAIN_THREAD_DURATION = "db_on_main_thread_duration_threshold"
     FILE_IO_MAIN_THREAD = "file_io_on_main_thread_detection_enabled"
@@ -103,6 +104,7 @@ class ConfigurableThresholds(Enum):
     ConfigurableThresholds.UNCOMPRESSED_ASSET_DURATION.value: ConfigurableThresholds.UNCOMPRESSED_ASSET.value,
     ConfigurableThresholds.UNCOMPRESSED_ASSET_SIZE.value: ConfigurableThresholds.UNCOMPRESSED_ASSET.value,
     ConfigurableThresholds.LARGE_HTTP_PAYLOAD_SIZE.value: ConfigurableThresholds.LARGE_HTTP_PAYLOAD.value,
+    ConfigurableThresholds.LARGE_HTTP_PAYLOAD_FILTERED_PATHS.value: ConfigurableThresholds.LARGE_HTTP_PAYLOAD.value,
     ConfigurableThresholds.DB_ON_MAIN_THREAD_DURATION.value: ConfigurableThresholds.DB_ON_MAIN_THREAD.value,
     ConfigurableThresholds.FILE_IO_MAIN_THREAD_DURATION.value: ConfigurableThresholds.FILE_IO_MAIN_THREAD.value,
     ConfigurableThresholds.CONSECUTIVE_DB_QUERIES_MIN_TIME_SAVED.value: ConfigurableThresholds.CONSECUTIVE_DB_QUERIES.value,
@@ -133,6 +135,11 @@ class ProjectPerformanceIssueSettingsSerializer(serializers.Serializer):
     large_http_payload_size_threshold = serializers.IntegerField(
         required=False, min_value=100000, max_value=TEN_MB
     )
+    large_http_payload_filtered_paths = serializers.CharField(
+        required=False,
+        allow_blank=True,
+        max_length=1000,
+    )
     db_on_main_thread_duration_threshold = serializers.IntegerField(
         required=False, min_value=10, max_value=50
     )

src/sentry/options/defaults.py

@@ -1878,6 +1878,11 @@
     default=300000,
     flags=FLAG_AUTOMATOR_MODIFIABLE,
 )  # 1MB
+register(
+    "performance.issues.large_http_payload.filtered_paths",
+    default="",
+    flags=FLAG_AUTOMATOR_MODIFIABLE,
+)
 register(
     "performance.issues.db_on_main_thread.total_spans_duration_threshold",
     default=16,