fastapi-observer 0.3.1__tar.gz → 0.4.0__tar.gz

{fastapi_observer-0.3.1/src/fastapi_observer.egg-info → fastapi_observer-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fastapi-observer
-Version: 0.3.1
+Version: 0.4.0
 Summary: Zero-glue FastAPI observability with security presets and runtime controls
 Author-email: Vitaee <opensource@vitaee.dev>
 License-Expression: MIT
@@ -65,6 +65,7 @@ Requires-Dist: pytest>=9.0.2; extra == "dev"
 Requires-Dist: pytest-randomly>=3.16.0; extra == "dev"
 Requires-Dist: ruff>=0.15.1; extra == "dev"
 Requires-Dist: cyclonedx-bom>=7.2.1; extra == "dev"
+Requires-Dist: uvicorn>=0.30.0; extra == "dev"
 Dynamic: license-file
 
 # fastapi-observer
@@ -391,6 +392,19 @@ Request arrives
   -> context cleared
 ```
 
+### Internal Package Layout (Contributor Map)
+
+The project is now organized as focused subpackages instead of large monolithic modules:
+
+- `fastapiobserver/logging/`: formatter, queueing, filters, setup lifecycle, sink circuit-breakers.
+- `fastapiobserver/middleware/`: request logging orchestration, context, IP resolution, headers, body capture, metrics hooks.
+- `fastapiobserver/sinks/`: sink protocol, registry/discovery, built-ins, factory wiring, Logtail + DLQ implementation.
+- `fastapiobserver/metrics/`: backend contracts/registry/builder/endpoint, Prometheus integration subpackage.
+- `fastapiobserver/security/`: policy/settings models, normalization helpers, redaction engine, trusted-proxy utilities.
+- `fastapiobserver/otel/`: OTel settings/resource/tracing/logs/metrics/lifecycle helpers.
+
+Public imports remain backward-compatible via package facades (`__init__.py` re-exports).
+
 ---
 
 ## Example JSON Log Event
@@ -570,11 +584,57 @@ kubectl -n observability rollout status deployment/traffic-generator
 kubectl -n observability port-forward svc/grafana 3000:3000
 ```
 
-Open [http://localhost:3000](http://localhost:3000).  
+Open [http://localhost:3000](http://localhost:3000).
 Full guide: [`kubernetes.md`](kubernetes.md)
 
 ---
 
+## Low-Overhead & Production Tuning (Advanced)
+
+`fastapi-observer` integrates natively with the core OpenTelemetry Python SDK, meaning you can aggressively tune its resource usage purely via standard environment variables without altering your application code.
+
+For high-throughput services (e.g. `10k+ RPS`), apply these exact variables to minimize the observer footprint:
+
+### 1. Head-Based Sampling
+
+Tracing 100% of requests is too expensive at scale. You should configure `fastapi-observer` to respect upstream trace flags, while only sampling a fraction of net-new requests:
+
+```bash
+# Keep the parent's sample decision if it exists, otherwise sample 5%
+export OTEL_TRACES_SAMPLER="parentbased_traceidratio"
+export OTEL_TRACES_SAMPLER_ARG="0.05"
+```
+
+### 2. Exclude Noisy URLs from the SDK
+
+Do not waste cycles generating spans for health checks or static assets. `fastapi-observer` will auto-derive metrics exclusions, but you can explicitly drop them from tracing at the C-extension level:
+
+```bash
+export OTEL_PYTHON_FASTAPI_EXCLUDED_URLS="healthz,metrics,favicon.ico"
+```
+
+### 3. Cap Span Attributes
+
+Prevent large, unmanageable spans from consuming excessive memory in the `BatchSpanProcessor`:
+
+```bash
+export OTEL_SPAN_ATTRIBUTE_COUNT_LIMIT="128"
+export OTEL_SPAN_EVENT_COUNT_LIMIT="128"
+export OTEL_SPAN_LINK_COUNT_LIMIT="128"
+```
+
+### 4. Optimize Output Buffers
+
+The default OpenTelemetry batch limits are too conservative for high-throughput ASGI microservices. Increase the max queue limits so spikes aren't dropped, but decrease the timeout so the process memory is flushed faster:
+
+```bash
+export OTEL_BSP_MAX_QUEUE_SIZE="10000"
+export OTEL_BSP_MAX_EXPORT_BATCH_SIZE="5000"
+export OTEL_BSP_SCHEDULE_DELAY="1000"
+```
+
+---
+
 ## Examples
 
 The `examples/` directory contains runnable demos:
@@ -729,7 +789,36 @@ install_observability(app, settings, security_policy=SecurityPolicy(log_request_
 app.add_middleware(CORSMiddleware, allow_origins=["*"])
 ```
 
-### Multi-worker Gunicorn with Prometheus
+### Multi-worker Gunicorn
+
+> [!WARNING]
+> `--preload-app` is dangerous if visibility is initialized at the module level.
+>
+> When `--preload-app` is used, the application is loaded into the master process *before* forking the worker processes. `fastapi-observer` uses a highly-performant background thread (`QueueListener`) for non-blocking logging, and OpenTelemetry similarly spawns background export threads.
+>
+> Threads **do not survive** a process fork. If you call `install_observability()` at the module level (e.g., right under `app = FastAPI()`), the background threads will be created in the master process, and all workers will silently drop logs because their logging threads never started.
+>
+> **How to fix:** Either remove `--preload-app`, OR initialize observability inside the FastAPI `lifespan` context manager so it starts safely after the fork inside each worker:
+>
+> ```python
+> from contextlib import asynccontextmanager
+> from fastapi import FastAPI
+> from fastapiobserver import ObservabilitySettings, install_observability
+>
+> settings = ObservabilitySettings(service="api")
+>
+> @asynccontextmanager
+> async def lifespan(app: FastAPI):
+>     install_observability(app, settings) # Safely initializes per-worker
+>     yield
+>     # fastapiobserver automatically registers shutdown hooks so no teardown needed here
+>
+> app = FastAPI(lifespan=lifespan)
+> ```
+
+#### Prometheus Multiprocess Mode
+
+If you are using Prometheus with multiple Gunicorn workers, you must configure a shared metrics directory:
 
 ```bash
 export PROMETHEUS_MULTIPROC_DIR=/tmp/prometheus-metrics
@@ -961,9 +1050,10 @@ Reproducible benchmark harness and methodology:
 
 - `0.1.x`: secure-by-default core
 - `0.2.x`: OTel interoperability, security presets, allowlists
+- `0.3.x`: package modularization, sink/metrics hardening, and performance reliability updates
 - `1.0.0`: dynamic runtime controls and plugin stability
 
-Current release version: `0.3.1`
+Current release version: `0.4.0`
 
 ## Changelog Policy
 
@@ -1018,4 +1108,4 @@ The pre-push hook runs:
 
 ## Roadmap Tracking
 
-See [NEXT_STEPS.md](NEXT_STEPS.md) for the active `0.2.0` roadmap and release checklist.
+See [NEXT_STEPS.md](NEXT_STEPS.md) for the active roadmap and release checklist.

{fastapi_observer-0.3.1 → fastapi_observer-0.4.0}/README.md

@@ -322,6 +322,19 @@ Request arrives
   -> context cleared
 ```
 
+### Internal Package Layout (Contributor Map)
+
+The project is now organized as focused subpackages instead of large monolithic modules:
+
+- `fastapiobserver/logging/`: formatter, queueing, filters, setup lifecycle, sink circuit-breakers.
+- `fastapiobserver/middleware/`: request logging orchestration, context, IP resolution, headers, body capture, metrics hooks.
+- `fastapiobserver/sinks/`: sink protocol, registry/discovery, built-ins, factory wiring, Logtail + DLQ implementation.
+- `fastapiobserver/metrics/`: backend contracts/registry/builder/endpoint, Prometheus integration subpackage.
+- `fastapiobserver/security/`: policy/settings models, normalization helpers, redaction engine, trusted-proxy utilities.
+- `fastapiobserver/otel/`: OTel settings/resource/tracing/logs/metrics/lifecycle helpers.
+
+Public imports remain backward-compatible via package facades (`__init__.py` re-exports).
+
 ---
 
 ## Example JSON Log Event
@@ -501,11 +514,57 @@ kubectl -n observability rollout status deployment/traffic-generator
 kubectl -n observability port-forward svc/grafana 3000:3000
 ```
 
-Open [http://localhost:3000](http://localhost:3000).  
+Open [http://localhost:3000](http://localhost:3000).
 Full guide: [`kubernetes.md`](kubernetes.md)
 
 ---
 
+## Low-Overhead & Production Tuning (Advanced)
+
+`fastapi-observer` integrates natively with the core OpenTelemetry Python SDK, meaning you can aggressively tune its resource usage purely via standard environment variables without altering your application code.
+
+For high-throughput services (e.g. `10k+ RPS`), apply these exact variables to minimize the observer footprint:
+
+### 1. Head-Based Sampling
+
+Tracing 100% of requests is too expensive at scale. You should configure `fastapi-observer` to respect upstream trace flags, while only sampling a fraction of net-new requests:
+
+```bash
+# Keep the parent's sample decision if it exists, otherwise sample 5%
+export OTEL_TRACES_SAMPLER="parentbased_traceidratio"
+export OTEL_TRACES_SAMPLER_ARG="0.05"
+```
+
+### 2. Exclude Noisy URLs from the SDK
+
+Do not waste cycles generating spans for health checks or static assets. `fastapi-observer` will auto-derive metrics exclusions, but you can explicitly drop them from tracing at the C-extension level:
+
+```bash
+export OTEL_PYTHON_FASTAPI_EXCLUDED_URLS="healthz,metrics,favicon.ico"
+```
+
+### 3. Cap Span Attributes
+
+Prevent large, unmanageable spans from consuming excessive memory in the `BatchSpanProcessor`:
+
+```bash
+export OTEL_SPAN_ATTRIBUTE_COUNT_LIMIT="128"
+export OTEL_SPAN_EVENT_COUNT_LIMIT="128"
+export OTEL_SPAN_LINK_COUNT_LIMIT="128"
+```
+
+### 4. Optimize Output Buffers
+
+The default OpenTelemetry batch limits are too conservative for high-throughput ASGI microservices. Increase the max queue limits so spikes aren't dropped, but decrease the timeout so the process memory is flushed faster:
+
+```bash
+export OTEL_BSP_MAX_QUEUE_SIZE="10000"
+export OTEL_BSP_MAX_EXPORT_BATCH_SIZE="5000"
+export OTEL_BSP_SCHEDULE_DELAY="1000"
+```
+
+---
+
 ## Examples
 
 The `examples/` directory contains runnable demos:
@@ -660,7 +719,36 @@ install_observability(app, settings, security_policy=SecurityPolicy(log_request_
 app.add_middleware(CORSMiddleware, allow_origins=["*"])
 ```
 
-### Multi-worker Gunicorn with Prometheus
+### Multi-worker Gunicorn
+
+> [!WARNING]
+> `--preload-app` is dangerous if visibility is initialized at the module level.
+>
+> When `--preload-app` is used, the application is loaded into the master process *before* forking the worker processes. `fastapi-observer` uses a highly-performant background thread (`QueueListener`) for non-blocking logging, and OpenTelemetry similarly spawns background export threads.
+>
+> Threads **do not survive** a process fork. If you call `install_observability()` at the module level (e.g., right under `app = FastAPI()`), the background threads will be created in the master process, and all workers will silently drop logs because their logging threads never started.
+>
+> **How to fix:** Either remove `--preload-app`, OR initialize observability inside the FastAPI `lifespan` context manager so it starts safely after the fork inside each worker:
+>
+> ```python
+> from contextlib import asynccontextmanager
+> from fastapi import FastAPI
+> from fastapiobserver import ObservabilitySettings, install_observability
+>
+> settings = ObservabilitySettings(service="api")
+>
+> @asynccontextmanager
+> async def lifespan(app: FastAPI):
+>     install_observability(app, settings) # Safely initializes per-worker
+>     yield
+>     # fastapiobserver automatically registers shutdown hooks so no teardown needed here
+>
+> app = FastAPI(lifespan=lifespan)
+> ```
+
+#### Prometheus Multiprocess Mode
+
+If you are using Prometheus with multiple Gunicorn workers, you must configure a shared metrics directory:
 
 ```bash
 export PROMETHEUS_MULTIPROC_DIR=/tmp/prometheus-metrics
@@ -892,9 +980,10 @@ Reproducible benchmark harness and methodology:
 
 - `0.1.x`: secure-by-default core
 - `0.2.x`: OTel interoperability, security presets, allowlists
+- `0.3.x`: package modularization, sink/metrics hardening, and performance reliability updates
 - `1.0.0`: dynamic runtime controls and plugin stability
 
-Current release version: `0.3.1`
+Current release version: `0.4.0`
 
 ## Changelog Policy
 
@@ -949,4 +1038,4 @@ The pre-push hook runs:
 
 ## Roadmap Tracking
 
-See [NEXT_STEPS.md](NEXT_STEPS.md) for the active `0.2.0` roadmap and release checklist.
+See [NEXT_STEPS.md](NEXT_STEPS.md) for the active roadmap and release checklist.

{fastapi_observer-0.3.1 → fastapi_observer-0.4.0}/pyproject.toml

@@ -86,6 +86,7 @@ dev = [
     "pytest-randomly>=3.16.0",
     "ruff>=0.15.1",
     "cyclonedx-bom>=7.2.1",
+    "uvicorn>=0.30.0",
 ]
 
 [tool.setuptools]

{fastapi_observer-0.3.1 → fastapi_observer-0.4.0/src/fastapi_observer.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fastapi-observer
-Version: 0.3.1
+Version: 0.4.0
 Summary: Zero-glue FastAPI observability with security presets and runtime controls
 Author-email: Vitaee <opensource@vitaee.dev>
 License-Expression: MIT
@@ -65,6 +65,7 @@ Requires-Dist: pytest>=9.0.2; extra == "dev"
 Requires-Dist: pytest-randomly>=3.16.0; extra == "dev"
 Requires-Dist: ruff>=0.15.1; extra == "dev"
 Requires-Dist: cyclonedx-bom>=7.2.1; extra == "dev"
+Requires-Dist: uvicorn>=0.30.0; extra == "dev"
 Dynamic: license-file
 
 # fastapi-observer
@@ -391,6 +392,19 @@ Request arrives
   -> context cleared
 ```
 
+### Internal Package Layout (Contributor Map)
+
+The project is now organized as focused subpackages instead of large monolithic modules:
+
+- `fastapiobserver/logging/`: formatter, queueing, filters, setup lifecycle, sink circuit-breakers.
+- `fastapiobserver/middleware/`: request logging orchestration, context, IP resolution, headers, body capture, metrics hooks.
+- `fastapiobserver/sinks/`: sink protocol, registry/discovery, built-ins, factory wiring, Logtail + DLQ implementation.
+- `fastapiobserver/metrics/`: backend contracts/registry/builder/endpoint, Prometheus integration subpackage.
+- `fastapiobserver/security/`: policy/settings models, normalization helpers, redaction engine, trusted-proxy utilities.
+- `fastapiobserver/otel/`: OTel settings/resource/tracing/logs/metrics/lifecycle helpers.
+
+Public imports remain backward-compatible via package facades (`__init__.py` re-exports).
+
 ---
 
 ## Example JSON Log Event
@@ -570,11 +584,57 @@ kubectl -n observability rollout status deployment/traffic-generator
 kubectl -n observability port-forward svc/grafana 3000:3000
 ```
 
-Open [http://localhost:3000](http://localhost:3000).  
+Open [http://localhost:3000](http://localhost:3000).
 Full guide: [`kubernetes.md`](kubernetes.md)
 
 ---
 
+## Low-Overhead & Production Tuning (Advanced)
+
+`fastapi-observer` integrates natively with the core OpenTelemetry Python SDK, meaning you can aggressively tune its resource usage purely via standard environment variables without altering your application code.
+
+For high-throughput services (e.g. `10k+ RPS`), apply these exact variables to minimize the observer footprint:
+
+### 1. Head-Based Sampling
+
+Tracing 100% of requests is too expensive at scale. You should configure `fastapi-observer` to respect upstream trace flags, while only sampling a fraction of net-new requests:
+
+```bash
+# Keep the parent's sample decision if it exists, otherwise sample 5%
+export OTEL_TRACES_SAMPLER="parentbased_traceidratio"
+export OTEL_TRACES_SAMPLER_ARG="0.05"
+```
+
+### 2. Exclude Noisy URLs from the SDK
+
+Do not waste cycles generating spans for health checks or static assets. `fastapi-observer` will auto-derive metrics exclusions, but you can explicitly drop them from tracing at the C-extension level:
+
+```bash
+export OTEL_PYTHON_FASTAPI_EXCLUDED_URLS="healthz,metrics,favicon.ico"
+```
+
+### 3. Cap Span Attributes
+
+Prevent large, unmanageable spans from consuming excessive memory in the `BatchSpanProcessor`:
+
+```bash
+export OTEL_SPAN_ATTRIBUTE_COUNT_LIMIT="128"
+export OTEL_SPAN_EVENT_COUNT_LIMIT="128"
+export OTEL_SPAN_LINK_COUNT_LIMIT="128"
+```
+
+### 4. Optimize Output Buffers
+
+The default OpenTelemetry batch limits are too conservative for high-throughput ASGI microservices. Increase the max queue limits so spikes aren't dropped, but decrease the timeout so the process memory is flushed faster:
+
+```bash
+export OTEL_BSP_MAX_QUEUE_SIZE="10000"
+export OTEL_BSP_MAX_EXPORT_BATCH_SIZE="5000"
+export OTEL_BSP_SCHEDULE_DELAY="1000"
+```
+
+---
+
 ## Examples
 
 The `examples/` directory contains runnable demos:
@@ -729,7 +789,36 @@ install_observability(app, settings, security_policy=SecurityPolicy(log_request_
 app.add_middleware(CORSMiddleware, allow_origins=["*"])
 ```
 
-### Multi-worker Gunicorn with Prometheus
+### Multi-worker Gunicorn
+
+> [!WARNING]
+> `--preload-app` is dangerous if visibility is initialized at the module level.
+>
+> When `--preload-app` is used, the application is loaded into the master process *before* forking the worker processes. `fastapi-observer` uses a highly-performant background thread (`QueueListener`) for non-blocking logging, and OpenTelemetry similarly spawns background export threads.
+>
+> Threads **do not survive** a process fork. If you call `install_observability()` at the module level (e.g., right under `app = FastAPI()`), the background threads will be created in the master process, and all workers will silently drop logs because their logging threads never started.
+>
+> **How to fix:** Either remove `--preload-app`, OR initialize observability inside the FastAPI `lifespan` context manager so it starts safely after the fork inside each worker:
+>
+> ```python
+> from contextlib import asynccontextmanager
+> from fastapi import FastAPI
+> from fastapiobserver import ObservabilitySettings, install_observability
+>
+> settings = ObservabilitySettings(service="api")
+>
+> @asynccontextmanager
+> async def lifespan(app: FastAPI):
+>     install_observability(app, settings) # Safely initializes per-worker
+>     yield
+>     # fastapiobserver automatically registers shutdown hooks so no teardown needed here
+>
+> app = FastAPI(lifespan=lifespan)
+> ```
+
+#### Prometheus Multiprocess Mode
+
+If you are using Prometheus with multiple Gunicorn workers, you must configure a shared metrics directory:
 
 ```bash
 export PROMETHEUS_MULTIPROC_DIR=/tmp/prometheus-metrics
@@ -961,9 +1050,10 @@ Reproducible benchmark harness and methodology:
 
 - `0.1.x`: secure-by-default core
 - `0.2.x`: OTel interoperability, security presets, allowlists
+- `0.3.x`: package modularization, sink/metrics hardening, and performance reliability updates
 - `1.0.0`: dynamic runtime controls and plugin stability
 
-Current release version: `0.3.1`
+Current release version: `0.4.0`
 
 ## Changelog Policy
 
@@ -1018,4 +1108,4 @@ The pre-push hook runs:
 
 ## Roadmap Tracking
 
-See [NEXT_STEPS.md](NEXT_STEPS.md) for the active `0.2.0` roadmap and release checklist.
+See [NEXT_STEPS.md](NEXT_STEPS.md) for the active roadmap and release checklist.

fastapi_observer-0.4.0/src/fastapi_observer.egg-info/SOURCES.txt

@@ -0,0 +1,104 @@
+LICENSE
+README.md
+pyproject.toml
+src/fastapi_observer.egg-info/PKG-INFO
+src/fastapi_observer.egg-info/SOURCES.txt
+src/fastapi_observer.egg-info/dependency_links.txt
+src/fastapi_observer.egg-info/requires.txt
+src/fastapi_observer.egg-info/top_level.txt
+src/fastapiobserver/__init__.py
+src/fastapiobserver/_version.py
+src/fastapiobserver/config.py
+src/fastapiobserver/control_plane.py
+src/fastapiobserver/fastapi.py
+src/fastapiobserver/loguru.py
+src/fastapiobserver/plugins.py
+src/fastapiobserver/propagation.py
+src/fastapiobserver/py.typed
+src/fastapiobserver/request_context.py
+src/fastapiobserver/utils.py
+src/fastapiobserver/integrations/__init__.py
+src/fastapiobserver/integrations/strawberry.py
+src/fastapiobserver/logging/__init__.py
+src/fastapiobserver/logging/circuit_breaker.py
+src/fastapiobserver/logging/filters.py
+src/fastapiobserver/logging/formatter.py
+src/fastapiobserver/logging/queueing.py
+src/fastapiobserver/logging/setup.py
+src/fastapiobserver/logging/state.py
+src/fastapiobserver/metrics/__init__.py
+src/fastapiobserver/metrics/builder.py
+src/fastapiobserver/metrics/contracts.py
+src/fastapiobserver/metrics/endpoint.py
+src/fastapiobserver/metrics/noop.py
+src/fastapiobserver/metrics/pathing.py
+src/fastapiobserver/metrics/registry.py
+src/fastapiobserver/metrics/prometheus/__init__.py
+src/fastapiobserver/metrics/prometheus/backend.py
+src/fastapiobserver/metrics/prometheus/client.py
+src/fastapiobserver/metrics/prometheus/collector.py
+src/fastapiobserver/metrics/prometheus/exemplars.py
+src/fastapiobserver/metrics/prometheus/multiprocess.py
+src/fastapiobserver/middleware/__init__.py
+src/fastapiobserver/middleware/body_capture.py
+src/fastapiobserver/middleware/context.py
+src/fastapiobserver/middleware/events.py
+src/fastapiobserver/middleware/headers.py
+src/fastapiobserver/middleware/ip.py
+src/fastapiobserver/middleware/metrics.py
+src/fastapiobserver/middleware/request_logging.py
+src/fastapiobserver/middleware/span_errors.py
+src/fastapiobserver/otel/__init__.py
+src/fastapiobserver/otel/lifecycle.py
+src/fastapiobserver/otel/logs.py
+src/fastapiobserver/otel/metrics.py
+src/fastapiobserver/otel/resource.py
+src/fastapiobserver/otel/settings.py
+src/fastapiobserver/otel/tracing.py
+src/fastapiobserver/otel/types.py
+src/fastapiobserver/security/__init__.py
+src/fastapiobserver/security/normalize.py
+src/fastapiobserver/security/policy.py
+src/fastapiobserver/security/proxies.py
+src/fastapiobserver/security/redaction.py
+src/fastapiobserver/sinks/__init__.py
+src/fastapiobserver/sinks/builtin.py
+src/fastapiobserver/sinks/discovery.py
+src/fastapiobserver/sinks/factory.py
+src/fastapiobserver/sinks/protocol.py
+src/fastapiobserver/sinks/registry.py
+src/fastapiobserver/sinks/stats.py
+src/fastapiobserver/sinks/logtail/__init__.py
+src/fastapiobserver/sinks/logtail/dlq.py
+src/fastapiobserver/sinks/logtail/handler.py
+src/fastapiobserver/sinks/logtail/sink.py
+tests/test_body_media_allowlist.py
+tests/test_config_settings.py
+tests/test_excluded_url_precedence.py
+tests/test_fastapi_install.py
+tests/test_graphql_integrations.py
+tests/test_log_queue_controls.py
+tests/test_logging_schema_contract.py
+tests/test_logtail_dlq.py
+tests/test_loguru_bridge.py
+tests/test_metrics_contracts.py
+tests/test_metrics_optional_dependency.py
+tests/test_middleware.py
+tests/test_negotiate_content_type.py
+tests/test_otel_integration.py
+tests/test_otel_lifecycle.py
+tests/test_otel_log_correlation.py
+tests/test_otel_logs_from_env.py
+tests/test_otel_logs_pipeline.py
+tests/test_otel_metrics_from_env.py
+tests/test_otlp_export_integration.py
+tests/test_plugin_isolation.py
+tests/test_propagation.py
+tests/test_redaction_presets.py
+tests/test_route_template_cardinality.py
+tests/test_runtime_control_auth.py
+tests/test_security_redaction.py
+tests/test_sink_circuit_breaker.py
+tests/test_sinks.py
+tests/test_trust_boundary.py
+tests/test_utils_refactors.py

{fastapi_observer-0.3.1 → fastapi_observer-0.4.0}/src/fastapi_observer.egg-info/requires.txt

@@ -23,6 +23,7 @@ pytest>=9.0.2
 pytest-randomly>=3.16.0
 ruff>=0.15.1
 cyclonedx-bom>=7.2.1
+uvicorn>=0.30.0
 
 [fast-json]
 orjson>=3.11.7

fastapi_observer-0.4.0/src/fastapiobserver/_version.py

@@ -0,0 +1 @@
+__version__ = "0.4.0"

{fastapi_observer-0.3.1 → fastapi_observer-0.4.0}/src/fastapiobserver/control_plane.py

@@ -72,7 +72,13 @@ def mount_control_plane(app: FastAPI, settings: RuntimeControlSettings) -> None:
     def authorize(
         authorization: Annotated[str | None, Header()] = None,
     ) -> None:
-        expected = f"Bearer {token}"
+        current_token = os.getenv(settings.token_env_var)
+        if not current_token:
+            raise HTTPException(
+                status_code=status.HTTP_401_UNAUTHORIZED,
+                detail="Unauthorized",
+            )
+        expected = f"Bearer {current_token}"
         if not authorization or not secrets.compare_digest(authorization, expected):
             raise HTTPException(
                 status_code=status.HTTP_401_UNAUTHORIZED,

fastapi_observer-0.4.0/src/fastapiobserver/logging/__init__.py

@@ -0,0 +1,36 @@
+from __future__ import annotations
+
+from .circuit_breaker import (
+    CircuitBreakerState,
+    SinkCircuitBreakerHandler,
+    SinkCircuitBreakerSnapshot,
+    get_sink_circuit_breaker_stats,
+)
+from .filters import PluginLogFilter, RequestIdFilter, TraceContextFilter
+from .formatter import LOG_SCHEMA_VERSION, StructuredJsonFormatter
+from .queueing import (
+    LogQueueStatsSnapshot,
+    LogQueueTelemetry,
+    OverflowPolicyQueueHandler,
+    get_log_queue_stats,
+)
+from .setup import get_managed_output_handlers, setup_logging, shutdown_logging
+
+__all__ = [
+    "LOG_SCHEMA_VERSION",
+    "CircuitBreakerState",
+    "LogQueueStatsSnapshot",
+    "LogQueueTelemetry",
+    "OverflowPolicyQueueHandler",
+    "PluginLogFilter",
+    "RequestIdFilter",
+    "SinkCircuitBreakerHandler",
+    "SinkCircuitBreakerSnapshot",
+    "StructuredJsonFormatter",
+    "TraceContextFilter",
+    "get_log_queue_stats",
+    "get_managed_output_handlers",
+    "get_sink_circuit_breaker_stats",
+    "setup_logging",
+    "shutdown_logging",
+]