fastx-opentelemetry 0.1.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- fastx_opentelemetry-0.1.0/.gitignore +22 -0
- fastx_opentelemetry-0.1.0/PKG-INFO +329 -0
- fastx_opentelemetry-0.1.0/README.md +297 -0
- fastx_opentelemetry-0.1.0/examples/app.py +130 -0
- fastx_opentelemetry-0.1.0/examples/basic_usage.py +6 -0
- fastx_opentelemetry-0.1.0/examples/fa.py +23 -0
- fastx_opentelemetry-0.1.0/examples/full_usage.py +34 -0
- fastx_opentelemetry-0.1.0/examples/genai_usage.py +26 -0
- fastx_opentelemetry-0.1.0/fastx_opentelemetry/__init__.py +273 -0
- fastx_opentelemetry-0.1.0/fastx_opentelemetry/_config.py +45 -0
- fastx_opentelemetry-0.1.0/fastx_opentelemetry/_instrumentor.py +143 -0
- fastx_opentelemetry-0.1.0/fastx_opentelemetry/_resource.py +99 -0
- fastx_opentelemetry-0.1.0/fastx_opentelemetry/_sampler.py +140 -0
- fastx_opentelemetry-0.1.0/fastx_opentelemetry/_span_processor.py +144 -0
- fastx_opentelemetry-0.1.0/pyproject.toml +59 -0
- fastx_opentelemetry-0.1.0/tests/conftest.py +28 -0
- fastx_opentelemetry-0.1.0/tests/test_config.py +89 -0
- fastx_opentelemetry-0.1.0/tests/test_init.py +292 -0
- fastx_opentelemetry-0.1.0/tests/test_instrumentor.py +261 -0
- fastx_opentelemetry-0.1.0/tests/test_resource.py +209 -0
- fastx_opentelemetry-0.1.0/tests/test_sampler.py +252 -0
- fastx_opentelemetry-0.1.0/tests/test_span_processor.py +207 -0
- fastx_opentelemetry-0.1.0/tox.ini +21 -0
|
@@ -0,0 +1,329 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: fastx-opentelemetry
|
|
3
|
+
Version: 0.1.0
|
|
4
|
+
Summary: Code-based OpenTelemetry auto-instrumentation for Python
|
|
5
|
+
Author-email: jimo <jimo@foxmail.com>
|
|
6
|
+
License: Apache-2.0
|
|
7
|
+
Keywords: auto-instrumentation,observability,opentelemetry,tracing
|
|
8
|
+
Classifier: Development Status :: 3 - Alpha
|
|
9
|
+
Classifier: Intended Audience :: Developers
|
|
10
|
+
Classifier: License :: OSI Approved :: Apache Software License
|
|
11
|
+
Classifier: Programming Language :: Python :: 3
|
|
12
|
+
Classifier: Programming Language :: Python :: 3.7
|
|
13
|
+
Classifier: Programming Language :: Python :: 3.8
|
|
14
|
+
Classifier: Programming Language :: Python :: 3.9
|
|
15
|
+
Classifier: Programming Language :: Python :: 3.10
|
|
16
|
+
Classifier: Programming Language :: Python :: 3.11
|
|
17
|
+
Classifier: Programming Language :: Python :: 3.12
|
|
18
|
+
Classifier: Topic :: System :: Monitoring
|
|
19
|
+
Requires-Python: >=3.7
|
|
20
|
+
Requires-Dist: importlib-metadata>=1.0; python_version < '3.8'
|
|
21
|
+
Requires-Dist: opentelemetry-api>=1.22
|
|
22
|
+
Requires-Dist: opentelemetry-instrumentation>=0.43
|
|
23
|
+
Requires-Dist: opentelemetry-sdk>=1.22
|
|
24
|
+
Provides-Extra: all
|
|
25
|
+
Requires-Dist: opentelemetry-exporter-otlp-proto-grpc>=1.22; extra == 'all'
|
|
26
|
+
Requires-Dist: opentelemetry-exporter-otlp-proto-http>=1.22; extra == 'all'
|
|
27
|
+
Provides-Extra: otlp-grpc
|
|
28
|
+
Requires-Dist: opentelemetry-exporter-otlp-proto-grpc>=1.22; extra == 'otlp-grpc'
|
|
29
|
+
Provides-Extra: otlp-http
|
|
30
|
+
Requires-Dist: opentelemetry-exporter-otlp-proto-http>=1.22; extra == 'otlp-http'
|
|
31
|
+
Description-Content-Type: text/markdown
|
|
32
|
+
|
|
33
|
+
# fastx-opentelemetry
|
|
34
|
+
|
|
35
|
+
基于代码的 OpenTelemetry 自动增强 SDK 封装。一行代码完成 OpenTelemetry 初始化,自动发现并加载所有已安装的 instrumentation 库。
|
|
36
|
+
|
|
37
|
+
## 特性
|
|
38
|
+
|
|
39
|
+
- **一行初始化** — `fastx_opentelemetry.init(service_name="my-app")` 完成全部配置
|
|
40
|
+
- **自动发现增强** — 自动检测并加载所有已安装的 `opentelemetry-instrumentation-*` 包
|
|
41
|
+
- **智能建议** — 检测已安装但未增强的库,提示安装对应 instrumentation 包
|
|
42
|
+
- **上下游服务串联** — 自动传播 Baggage 中的服务调用链(`parent.service.name` / `parent.service.names` / `current.service.name`)
|
|
43
|
+
- **Baggage → Span 属性** — 自动将 Baggage 条目复制为 Span 属性,无需手动操作
|
|
44
|
+
- **内置采样器** — 提供 `RulesBasedSampler`(按规则采样)和 `RateLimitingSampler`(限流采样)
|
|
45
|
+
- **自动资源检测** — 自动采集 hostname、IP、PID、instance ID 以及云平台属性(`cloud.platform`、`cloud.region` 等)
|
|
46
|
+
- **环境变量兼容** — 完全兼容 OpenTelemetry 标准环境变量
|
|
47
|
+
- **信号级端点** — 自动为 traces/metrics/logs 拼接 `/v1/traces`、`/v1/metrics`、`/v1/logs` 路径
|
|
48
|
+
|
|
49
|
+
## 安装
|
|
50
|
+
|
|
51
|
+
```bash
|
|
52
|
+
# 基础安装
|
|
53
|
+
pip install fastx-opentelemetry
|
|
54
|
+
|
|
55
|
+
# 带 OTLP gRPC 导出器
|
|
56
|
+
pip install fastx-opentelemetry[otlp-grpc]
|
|
57
|
+
|
|
58
|
+
# 带 OTLP HTTP 导出器
|
|
59
|
+
pip install fastx-opentelemetry[otlp-http]
|
|
60
|
+
|
|
61
|
+
# 全部安装
|
|
62
|
+
pip install fastx-opentelemetry[all]
|
|
63
|
+
```
|
|
64
|
+
|
|
65
|
+
## 快速开始
|
|
66
|
+
|
|
67
|
+
### 最简用法
|
|
68
|
+
|
|
69
|
+
```python
|
|
70
|
+
import fastx_opentelemetry
|
|
71
|
+
|
|
72
|
+
fastx_opentelemetry.init()
|
|
73
|
+
```
|
|
74
|
+
|
|
75
|
+
配合环境变量:
|
|
76
|
+
|
|
77
|
+
```bash
|
|
78
|
+
export OTEL_EXPORTER_OTLP_ENDPOINT=http://collector:4318
|
|
79
|
+
export OTEL_SERVICE_NAME=my-app
|
|
80
|
+
python my_app.py
|
|
81
|
+
```
|
|
82
|
+
|
|
83
|
+
### Flask 应用
|
|
84
|
+
|
|
85
|
+
```python
|
|
86
|
+
import fastx_opentelemetry
|
|
87
|
+
|
|
88
|
+
# 必须在创建 Flask app 之前初始化
|
|
89
|
+
fastx_opentelemetry.init(service_name="my-flask-app")
|
|
90
|
+
|
|
91
|
+
from flask import Flask
|
|
92
|
+
app = Flask(__name__)
|
|
93
|
+
|
|
94
|
+
@app.route("/hello")
|
|
95
|
+
def hello():
|
|
96
|
+
return {"message": "Hello, World!"}
|
|
97
|
+
```
|
|
98
|
+
|
|
99
|
+
### Django 应用
|
|
100
|
+
|
|
101
|
+
在 `settings.py` 中添加:
|
|
102
|
+
|
|
103
|
+
```python
|
|
104
|
+
import fastx_opentelemetry
|
|
105
|
+
fastx_opentelemetry.init(service_name="my-django-app")
|
|
106
|
+
```
|
|
107
|
+
|
|
108
|
+
### FastAPI 应用
|
|
109
|
+
|
|
110
|
+
```python
|
|
111
|
+
import fastx_opentelemetry
|
|
112
|
+
|
|
113
|
+
fastx_opentelemetry.init(service_name="my-fastapi-app")
|
|
114
|
+
|
|
115
|
+
from fastapi import FastAPI
|
|
116
|
+
app = FastAPI()
|
|
117
|
+
|
|
118
|
+
@app.get("/hello")
|
|
119
|
+
async def hello():
|
|
120
|
+
return {"message": "Hello, World!"}
|
|
121
|
+
```
|
|
122
|
+
|
|
123
|
+
## 高级用法
|
|
124
|
+
|
|
125
|
+
### 自定义采样器
|
|
126
|
+
|
|
127
|
+
```python
|
|
128
|
+
import fastx_opentelemetry
|
|
129
|
+
from fastx_opentelemetry import RulesBasedSampler, SamplingRule, RateLimitingSampler
|
|
130
|
+
from opentelemetry.sdk.trace.sampling import ALWAYS_ON, ALWAYS_OFF, TraceIdRatioBased
|
|
131
|
+
|
|
132
|
+
sampler = RulesBasedSampler(
|
|
133
|
+
rules=[
|
|
134
|
+
SamplingRule(match_name="/health.*", sampler=ALWAYS_ON),
|
|
135
|
+
SamplingRule(match_name="/static/.*", sampler=ALWAYS_OFF),
|
|
136
|
+
SamplingRule(match_name="/api/.*", sampler=TraceIdRatioBased(0.1)),
|
|
137
|
+
],
|
|
138
|
+
default_sampler=RateLimitingSampler(max_spans_per_second=50),
|
|
139
|
+
)
|
|
140
|
+
|
|
141
|
+
fastx_opentelemetry.init(
|
|
142
|
+
service_name="my-app",
|
|
143
|
+
sampler=sampler,
|
|
144
|
+
)
|
|
145
|
+
```
|
|
146
|
+
|
|
147
|
+
### 完整配置
|
|
148
|
+
|
|
149
|
+
```python
|
|
150
|
+
import fastx_opentelemetry
|
|
151
|
+
from opentelemetry.sdk.trace.export import ConsoleSpanExporter, SimpleSpanProcessor
|
|
152
|
+
|
|
153
|
+
console_processor = SimpleSpanProcessor(ConsoleSpanExporter())
|
|
154
|
+
|
|
155
|
+
loaded = fastx_opentelemetry.init(
|
|
156
|
+
service_name="my-app",
|
|
157
|
+
endpoint="http://collector:4317", # 自动拼接 /v1/traces, /v1/metrics, /v1/logs
|
|
158
|
+
protocol="grpc",
|
|
159
|
+
compression="gzip",
|
|
160
|
+
headers={"Authorization": "Bearer token123"},
|
|
161
|
+
extra_resource_attributes={
|
|
162
|
+
"deployment.environment": "production",
|
|
163
|
+
"service.version": "1.2.3",
|
|
164
|
+
},
|
|
165
|
+
span_processors=[console_processor],
|
|
166
|
+
disabled_instrumentations=["sqlite3"],
|
|
167
|
+
)
|
|
168
|
+
|
|
169
|
+
print(f"Loaded instrumentors: {loaded}")
|
|
170
|
+
```
|
|
171
|
+
|
|
172
|
+
### 禁用特定增强
|
|
173
|
+
|
|
174
|
+
```python
|
|
175
|
+
fastx_opentelemetry.init(
|
|
176
|
+
service_name="my-app",
|
|
177
|
+
disabled_instrumentations=["sqlite3", "urllib"],
|
|
178
|
+
)
|
|
179
|
+
```
|
|
180
|
+
|
|
181
|
+
或通过环境变量:
|
|
182
|
+
|
|
183
|
+
```bash
|
|
184
|
+
export OTEL_PYTHON_DISABLED_INSTRUMENTATIONS=sqlite3,urllib
|
|
185
|
+
```
|
|
186
|
+
|
|
187
|
+
禁用所有增强(仅初始化 SDK,不加载任何 instrumentor):
|
|
188
|
+
|
|
189
|
+
```python
|
|
190
|
+
fastx_opentelemetry.init(
|
|
191
|
+
service_name="my-app",
|
|
192
|
+
disabled_instrumentations=["*"],
|
|
193
|
+
)
|
|
194
|
+
```
|
|
195
|
+
|
|
196
|
+
### 上下游服务串联
|
|
197
|
+
|
|
198
|
+
`FastxBaggagePropagator` 和 `FastxSpanProcessor` 在 `init()` 时自动启用,无需额外配置。它们实现了与 Go 端 `fastx_otel.go` 一致的 Baggage 传播逻辑:
|
|
199
|
+
|
|
200
|
+
- **`FastxBaggagePropagator`** — 包装 W3C Baggage 传播器,提取时自动转换服务链:
|
|
201
|
+
- 将 `current.service.name` 移入 `parent.service.name`
|
|
202
|
+
- 将 `current.service.name` 追加到 `parent.service.names`(以 `++` 分隔,去重)
|
|
203
|
+
- 将 `current.service.name` 设置为本地服务名
|
|
204
|
+
- **`FastxSpanProcessor`** — 在 Span 开始时,自动将所有 Baggage 条目复制为 Span 属性
|
|
205
|
+
|
|
206
|
+
这意味着在微服务调用链中,每个服务的 Span 都会携带完整的上游服务链信息,便于在后端(如 Jaeger、Grafana)中追踪调用路径。
|
|
207
|
+
|
|
208
|
+
## 环境变量
|
|
209
|
+
|
|
210
|
+
| 变量 | 说明 | 默认值 |
|
|
211
|
+
|------|------|--------|
|
|
212
|
+
| `FASTX_PYTHON_INSTRUMENTATION_ENABLED` | 全局开关,`false` 时跳过全部初始化 | `true` |
|
|
213
|
+
| `OTEL_SERVICE_NAME` | 服务名称 | `unknown_service` |
|
|
214
|
+
| `OTEL_EXPORTER_OTLP_ENDPOINT` | OTLP Collector 地址(自动拼接信号路径) | — |
|
|
215
|
+
| `OTEL_EXPORTER_OTLP_PROTOCOL` | 传输协议 `http/protobuf` 或 `grpc` | `http/protobuf` |
|
|
216
|
+
| `OTEL_EXPORTER_OTLP_COMPRESSION` | 压缩方式 `gzip`/`deflate`/`none` | `gzip` |
|
|
217
|
+
| `OTEL_EXPORTER_OTLP_TRACES_HEADERS` | 导出器请求头 | — |
|
|
218
|
+
| `OTEL_TRACES_SAMPLER` | 采样器名称 | `parentbased_always_on` |
|
|
219
|
+
| `OTEL_PYTHON_DISABLED_INSTRUMENTATIONS` | 禁用的增强列表(逗号分隔,`*` 禁用全部) | — |
|
|
220
|
+
| `OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED` | 是否启用日志自动增强 | `true` |
|
|
221
|
+
| `IP_ADDR` | 自动检测为 `host.ip` 资源属性 | 本地非回环 IP |
|
|
222
|
+
| `CLOUD_TYPE` | 自动检测为 `cloud.platform` 资源属性 | — |
|
|
223
|
+
| `REGION_CODE` | 自动检测为 `cloud.region` 资源属性 | — |
|
|
224
|
+
| `AZ_CODE` | 自动检测为 `cloud.availability` 资源属性 | — |
|
|
225
|
+
| `env_cd` | 自动检测为 `deployment.environment.name` 资源属性 | — |
|
|
226
|
+
|
|
227
|
+
## 安装增强库
|
|
228
|
+
|
|
229
|
+
`fastx-opentelemetry` 会自动检测已安装的库并加载对应的 instrumentation。你需要手动安装需要的 instrumentation 包:
|
|
230
|
+
|
|
231
|
+
```bash
|
|
232
|
+
# Web 框架
|
|
233
|
+
pip install opentelemetry-instrumentation-flask
|
|
234
|
+
pip install opentelemetry-instrumentation-django
|
|
235
|
+
pip install opentelemetry-instrumentation-fastapi
|
|
236
|
+
|
|
237
|
+
# HTTP 客户端
|
|
238
|
+
pip install opentelemetry-instrumentation-requests
|
|
239
|
+
pip install opentelemetry-instrumentation-httpx
|
|
240
|
+
pip install opentelemetry-instrumentation-urllib
|
|
241
|
+
|
|
242
|
+
# 数据库
|
|
243
|
+
pip install opentelemetry-instrumentation-sqlalchemy
|
|
244
|
+
pip install opentelemetry-instrumentation-psycopg2
|
|
245
|
+
pip install opentelemetry-instrumentation-redis
|
|
246
|
+
pip install opentelemetry-instrumentation-pymongo
|
|
247
|
+
|
|
248
|
+
# 消息队列
|
|
249
|
+
pip install opentelemetry-instrumentation-celery
|
|
250
|
+
pip install opentelemetry-instrumentation-kafka-python
|
|
251
|
+
```
|
|
252
|
+
|
|
253
|
+
检查已安装的增强库:
|
|
254
|
+
|
|
255
|
+
```bash
|
|
256
|
+
pip list | grep opentelemetry-instrumentation
|
|
257
|
+
```
|
|
258
|
+
|
|
259
|
+
## API 参考
|
|
260
|
+
|
|
261
|
+
### `fastx_opentelemetry.init(**kwargs)`
|
|
262
|
+
|
|
263
|
+
初始化 OpenTelemetry SDK 并加载所有已安装的 instrumentor。
|
|
264
|
+
|
|
265
|
+
**参数:**
|
|
266
|
+
|
|
267
|
+
| 参数 | 类型 | 说明 |
|
|
268
|
+
|------|------|------|
|
|
269
|
+
| `service_name` | `str \| None` | 服务名称,回退到 `OTEL_SERVICE_NAME` |
|
|
270
|
+
| `endpoint` | `str \| None` | OTLP Collector 地址(自动拼接信号路径) |
|
|
271
|
+
| `protocol` | `str \| None` | `http/protobuf` 或 `grpc` |
|
|
272
|
+
| `compression` | `str \| None` | `gzip`、`deflate` 或 `none` |
|
|
273
|
+
| `headers` | `dict \| None` | OTLP 导出器请求头 |
|
|
274
|
+
| `extra_resource_attributes` | `dict \| None` | 额外的资源属性 |
|
|
275
|
+
| `span_processors` | `list \| None` | 自定义 SpanProcessor 列表 |
|
|
276
|
+
| `sampler` | `Sampler \| None` | 自定义采样器实例 |
|
|
277
|
+
| `disabled_instrumentations` | `list \| None` | 禁用的 instrumentor 名称列表,`"*"` 禁用全部 |
|
|
278
|
+
|
|
279
|
+
**返回值:** `list[str]` — 成功加载的 instrumentor 名称列表。
|
|
280
|
+
|
|
281
|
+
### `fastx_opentelemetry.list_available_instrumentors()`
|
|
282
|
+
|
|
283
|
+
返回所有已安装的 instrumentor 名称列表。
|
|
284
|
+
|
|
285
|
+
### `fastx_opentelemetry.suggest_missing_instrumentations(loaded)`
|
|
286
|
+
|
|
287
|
+
检查已安装但未增强的库,返回建议安装的 instrumentation 包名列表。
|
|
288
|
+
|
|
289
|
+
### 采样器
|
|
290
|
+
|
|
291
|
+
#### `RulesBasedSampler(rules, default_sampler=None)`
|
|
292
|
+
|
|
293
|
+
按规则匹配 span 名称或属性,委托给不同的采样器。
|
|
294
|
+
|
|
295
|
+
#### `SamplingRule(sampler, match_name=None, match_attributes=None)`
|
|
296
|
+
|
|
297
|
+
单条采样规则。`match_name` 为正则表达式匹配 span 名称,`match_attributes` 为属性键值对匹配。
|
|
298
|
+
|
|
299
|
+
#### `RateLimitingSampler(max_spans_per_second)`
|
|
300
|
+
|
|
301
|
+
基于令牌桶算法的限流采样器,限制每秒采样的 span 数量。
|
|
302
|
+
|
|
303
|
+
### Baggage 传播与 Span 处理
|
|
304
|
+
|
|
305
|
+
#### `FastxBaggagePropagator`
|
|
306
|
+
|
|
307
|
+
包装 W3C Baggage 传播器,在 `extract` 时自动转换服务调用链。`init()` 时自动设置为全局传播器。
|
|
308
|
+
|
|
309
|
+
#### `FastxSpanProcessor`
|
|
310
|
+
|
|
311
|
+
在 Span `on_start` 时将所有 Baggage 条目复制为 Span 属性。`init()` 时自动添加到 TracerProvider。
|
|
312
|
+
|
|
313
|
+
## 本地开发
|
|
314
|
+
|
|
315
|
+
```bash
|
|
316
|
+
# 安装依赖
|
|
317
|
+
pip install -e ".[all]"
|
|
318
|
+
|
|
319
|
+
# 运行示例
|
|
320
|
+
export OTEL_LOG_LEVEL=debug
|
|
321
|
+
export OTEL_EXPORTER_OTLP_ENDPOINT=http://127.0.0.1:4318
|
|
322
|
+
export OTEL_SERVICE_NAME=my-app
|
|
323
|
+
|
|
324
|
+
python examples/app.py
|
|
325
|
+
```
|
|
326
|
+
|
|
327
|
+
## License
|
|
328
|
+
|
|
329
|
+
Apache-2.0
|
|
@@ -0,0 +1,297 @@
|
|
|
1
|
+
# fastx-opentelemetry
|
|
2
|
+
|
|
3
|
+
基于代码的 OpenTelemetry 自动增强 SDK 封装。一行代码完成 OpenTelemetry 初始化,自动发现并加载所有已安装的 instrumentation 库。
|
|
4
|
+
|
|
5
|
+
## 特性
|
|
6
|
+
|
|
7
|
+
- **一行初始化** — `fastx_opentelemetry.init(service_name="my-app")` 完成全部配置
|
|
8
|
+
- **自动发现增强** — 自动检测并加载所有已安装的 `opentelemetry-instrumentation-*` 包
|
|
9
|
+
- **智能建议** — 检测已安装但未增强的库,提示安装对应 instrumentation 包
|
|
10
|
+
- **上下游服务串联** — 自动传播 Baggage 中的服务调用链(`parent.service.name` / `parent.service.names` / `current.service.name`)
|
|
11
|
+
- **Baggage → Span 属性** — 自动将 Baggage 条目复制为 Span 属性,无需手动操作
|
|
12
|
+
- **内置采样器** — 提供 `RulesBasedSampler`(按规则采样)和 `RateLimitingSampler`(限流采样)
|
|
13
|
+
- **自动资源检测** — 自动采集 hostname、IP、PID、instance ID 以及云平台属性(`cloud.platform`、`cloud.region` 等)
|
|
14
|
+
- **环境变量兼容** — 完全兼容 OpenTelemetry 标准环境变量
|
|
15
|
+
- **信号级端点** — 自动为 traces/metrics/logs 拼接 `/v1/traces`、`/v1/metrics`、`/v1/logs` 路径
|
|
16
|
+
|
|
17
|
+
## 安装
|
|
18
|
+
|
|
19
|
+
```bash
|
|
20
|
+
# 基础安装
|
|
21
|
+
pip install fastx-opentelemetry
|
|
22
|
+
|
|
23
|
+
# 带 OTLP gRPC 导出器
|
|
24
|
+
pip install fastx-opentelemetry[otlp-grpc]
|
|
25
|
+
|
|
26
|
+
# 带 OTLP HTTP 导出器
|
|
27
|
+
pip install fastx-opentelemetry[otlp-http]
|
|
28
|
+
|
|
29
|
+
# 全部安装
|
|
30
|
+
pip install fastx-opentelemetry[all]
|
|
31
|
+
```
|
|
32
|
+
|
|
33
|
+
## 快速开始
|
|
34
|
+
|
|
35
|
+
### 最简用法
|
|
36
|
+
|
|
37
|
+
```python
|
|
38
|
+
import fastx_opentelemetry
|
|
39
|
+
|
|
40
|
+
fastx_opentelemetry.init()
|
|
41
|
+
```
|
|
42
|
+
|
|
43
|
+
配合环境变量:
|
|
44
|
+
|
|
45
|
+
```bash
|
|
46
|
+
export OTEL_EXPORTER_OTLP_ENDPOINT=http://collector:4318
|
|
47
|
+
export OTEL_SERVICE_NAME=my-app
|
|
48
|
+
python my_app.py
|
|
49
|
+
```
|
|
50
|
+
|
|
51
|
+
### Flask 应用
|
|
52
|
+
|
|
53
|
+
```python
|
|
54
|
+
import fastx_opentelemetry
|
|
55
|
+
|
|
56
|
+
# 必须在创建 Flask app 之前初始化
|
|
57
|
+
fastx_opentelemetry.init(service_name="my-flask-app")
|
|
58
|
+
|
|
59
|
+
from flask import Flask
|
|
60
|
+
app = Flask(__name__)
|
|
61
|
+
|
|
62
|
+
@app.route("/hello")
|
|
63
|
+
def hello():
|
|
64
|
+
return {"message": "Hello, World!"}
|
|
65
|
+
```
|
|
66
|
+
|
|
67
|
+
### Django 应用
|
|
68
|
+
|
|
69
|
+
在 `settings.py` 中添加:
|
|
70
|
+
|
|
71
|
+
```python
|
|
72
|
+
import fastx_opentelemetry
|
|
73
|
+
fastx_opentelemetry.init(service_name="my-django-app")
|
|
74
|
+
```
|
|
75
|
+
|
|
76
|
+
### FastAPI 应用
|
|
77
|
+
|
|
78
|
+
```python
|
|
79
|
+
import fastx_opentelemetry
|
|
80
|
+
|
|
81
|
+
fastx_opentelemetry.init(service_name="my-fastapi-app")
|
|
82
|
+
|
|
83
|
+
from fastapi import FastAPI
|
|
84
|
+
app = FastAPI()
|
|
85
|
+
|
|
86
|
+
@app.get("/hello")
|
|
87
|
+
async def hello():
|
|
88
|
+
return {"message": "Hello, World!"}
|
|
89
|
+
```
|
|
90
|
+
|
|
91
|
+
## 高级用法
|
|
92
|
+
|
|
93
|
+
### 自定义采样器
|
|
94
|
+
|
|
95
|
+
```python
|
|
96
|
+
import fastx_opentelemetry
|
|
97
|
+
from fastx_opentelemetry import RulesBasedSampler, SamplingRule, RateLimitingSampler
|
|
98
|
+
from opentelemetry.sdk.trace.sampling import ALWAYS_ON, ALWAYS_OFF, TraceIdRatioBased
|
|
99
|
+
|
|
100
|
+
sampler = RulesBasedSampler(
|
|
101
|
+
rules=[
|
|
102
|
+
SamplingRule(match_name="/health.*", sampler=ALWAYS_ON),
|
|
103
|
+
SamplingRule(match_name="/static/.*", sampler=ALWAYS_OFF),
|
|
104
|
+
SamplingRule(match_name="/api/.*", sampler=TraceIdRatioBased(0.1)),
|
|
105
|
+
],
|
|
106
|
+
default_sampler=RateLimitingSampler(max_spans_per_second=50),
|
|
107
|
+
)
|
|
108
|
+
|
|
109
|
+
fastx_opentelemetry.init(
|
|
110
|
+
service_name="my-app",
|
|
111
|
+
sampler=sampler,
|
|
112
|
+
)
|
|
113
|
+
```
|
|
114
|
+
|
|
115
|
+
### 完整配置
|
|
116
|
+
|
|
117
|
+
```python
|
|
118
|
+
import fastx_opentelemetry
|
|
119
|
+
from opentelemetry.sdk.trace.export import ConsoleSpanExporter, SimpleSpanProcessor
|
|
120
|
+
|
|
121
|
+
console_processor = SimpleSpanProcessor(ConsoleSpanExporter())
|
|
122
|
+
|
|
123
|
+
loaded = fastx_opentelemetry.init(
|
|
124
|
+
service_name="my-app",
|
|
125
|
+
endpoint="http://collector:4317", # 自动拼接 /v1/traces, /v1/metrics, /v1/logs
|
|
126
|
+
protocol="grpc",
|
|
127
|
+
compression="gzip",
|
|
128
|
+
headers={"Authorization": "Bearer token123"},
|
|
129
|
+
extra_resource_attributes={
|
|
130
|
+
"deployment.environment": "production",
|
|
131
|
+
"service.version": "1.2.3",
|
|
132
|
+
},
|
|
133
|
+
span_processors=[console_processor],
|
|
134
|
+
disabled_instrumentations=["sqlite3"],
|
|
135
|
+
)
|
|
136
|
+
|
|
137
|
+
print(f"Loaded instrumentors: {loaded}")
|
|
138
|
+
```
|
|
139
|
+
|
|
140
|
+
### 禁用特定增强
|
|
141
|
+
|
|
142
|
+
```python
|
|
143
|
+
fastx_opentelemetry.init(
|
|
144
|
+
service_name="my-app",
|
|
145
|
+
disabled_instrumentations=["sqlite3", "urllib"],
|
|
146
|
+
)
|
|
147
|
+
```
|
|
148
|
+
|
|
149
|
+
或通过环境变量:
|
|
150
|
+
|
|
151
|
+
```bash
|
|
152
|
+
export OTEL_PYTHON_DISABLED_INSTRUMENTATIONS=sqlite3,urllib
|
|
153
|
+
```
|
|
154
|
+
|
|
155
|
+
禁用所有增强(仅初始化 SDK,不加载任何 instrumentor):
|
|
156
|
+
|
|
157
|
+
```python
|
|
158
|
+
fastx_opentelemetry.init(
|
|
159
|
+
service_name="my-app",
|
|
160
|
+
disabled_instrumentations=["*"],
|
|
161
|
+
)
|
|
162
|
+
```
|
|
163
|
+
|
|
164
|
+
### 上下游服务串联
|
|
165
|
+
|
|
166
|
+
`FastxBaggagePropagator` 和 `FastxSpanProcessor` 在 `init()` 时自动启用,无需额外配置。它们实现了与 Go 端 `fastx_otel.go` 一致的 Baggage 传播逻辑:
|
|
167
|
+
|
|
168
|
+
- **`FastxBaggagePropagator`** — 包装 W3C Baggage 传播器,提取时自动转换服务链:
|
|
169
|
+
- 将 `current.service.name` 移入 `parent.service.name`
|
|
170
|
+
- 将 `current.service.name` 追加到 `parent.service.names`(以 `++` 分隔,去重)
|
|
171
|
+
- 将 `current.service.name` 设置为本地服务名
|
|
172
|
+
- **`FastxSpanProcessor`** — 在 Span 开始时,自动将所有 Baggage 条目复制为 Span 属性
|
|
173
|
+
|
|
174
|
+
这意味着在微服务调用链中,每个服务的 Span 都会携带完整的上游服务链信息,便于在后端(如 Jaeger、Grafana)中追踪调用路径。
|
|
175
|
+
|
|
176
|
+
## 环境变量
|
|
177
|
+
|
|
178
|
+
| 变量 | 说明 | 默认值 |
|
|
179
|
+
|------|------|--------|
|
|
180
|
+
| `FASTX_PYTHON_INSTRUMENTATION_ENABLED` | 全局开关,`false` 时跳过全部初始化 | `true` |
|
|
181
|
+
| `OTEL_SERVICE_NAME` | 服务名称 | `unknown_service` |
|
|
182
|
+
| `OTEL_EXPORTER_OTLP_ENDPOINT` | OTLP Collector 地址(自动拼接信号路径) | — |
|
|
183
|
+
| `OTEL_EXPORTER_OTLP_PROTOCOL` | 传输协议 `http/protobuf` 或 `grpc` | `http/protobuf` |
|
|
184
|
+
| `OTEL_EXPORTER_OTLP_COMPRESSION` | 压缩方式 `gzip`/`deflate`/`none` | `gzip` |
|
|
185
|
+
| `OTEL_EXPORTER_OTLP_TRACES_HEADERS` | 导出器请求头 | — |
|
|
186
|
+
| `OTEL_TRACES_SAMPLER` | 采样器名称 | `parentbased_always_on` |
|
|
187
|
+
| `OTEL_PYTHON_DISABLED_INSTRUMENTATIONS` | 禁用的增强列表(逗号分隔,`*` 禁用全部) | — |
|
|
188
|
+
| `OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED` | 是否启用日志自动增强 | `true` |
|
|
189
|
+
| `IP_ADDR` | 自动检测为 `host.ip` 资源属性 | 本地非回环 IP |
|
|
190
|
+
| `CLOUD_TYPE` | 自动检测为 `cloud.platform` 资源属性 | — |
|
|
191
|
+
| `REGION_CODE` | 自动检测为 `cloud.region` 资源属性 | — |
|
|
192
|
+
| `AZ_CODE` | 自动检测为 `cloud.availability` 资源属性 | — |
|
|
193
|
+
| `env_cd` | 自动检测为 `deployment.environment.name` 资源属性 | — |
|
|
194
|
+
|
|
195
|
+
## 安装增强库
|
|
196
|
+
|
|
197
|
+
`fastx-opentelemetry` 会自动检测已安装的库并加载对应的 instrumentation。你需要手动安装需要的 instrumentation 包:
|
|
198
|
+
|
|
199
|
+
```bash
|
|
200
|
+
# Web 框架
|
|
201
|
+
pip install opentelemetry-instrumentation-flask
|
|
202
|
+
pip install opentelemetry-instrumentation-django
|
|
203
|
+
pip install opentelemetry-instrumentation-fastapi
|
|
204
|
+
|
|
205
|
+
# HTTP 客户端
|
|
206
|
+
pip install opentelemetry-instrumentation-requests
|
|
207
|
+
pip install opentelemetry-instrumentation-httpx
|
|
208
|
+
pip install opentelemetry-instrumentation-urllib
|
|
209
|
+
|
|
210
|
+
# 数据库
|
|
211
|
+
pip install opentelemetry-instrumentation-sqlalchemy
|
|
212
|
+
pip install opentelemetry-instrumentation-psycopg2
|
|
213
|
+
pip install opentelemetry-instrumentation-redis
|
|
214
|
+
pip install opentelemetry-instrumentation-pymongo
|
|
215
|
+
|
|
216
|
+
# 消息队列
|
|
217
|
+
pip install opentelemetry-instrumentation-celery
|
|
218
|
+
pip install opentelemetry-instrumentation-kafka-python
|
|
219
|
+
```
|
|
220
|
+
|
|
221
|
+
检查已安装的增强库:
|
|
222
|
+
|
|
223
|
+
```bash
|
|
224
|
+
pip list | grep opentelemetry-instrumentation
|
|
225
|
+
```
|
|
226
|
+
|
|
227
|
+
## API 参考
|
|
228
|
+
|
|
229
|
+
### `fastx_opentelemetry.init(**kwargs)`
|
|
230
|
+
|
|
231
|
+
初始化 OpenTelemetry SDK 并加载所有已安装的 instrumentor。
|
|
232
|
+
|
|
233
|
+
**参数:**
|
|
234
|
+
|
|
235
|
+
| 参数 | 类型 | 说明 |
|
|
236
|
+
|------|------|------|
|
|
237
|
+
| `service_name` | `str \| None` | 服务名称,回退到 `OTEL_SERVICE_NAME` |
|
|
238
|
+
| `endpoint` | `str \| None` | OTLP Collector 地址(自动拼接信号路径) |
|
|
239
|
+
| `protocol` | `str \| None` | `http/protobuf` 或 `grpc` |
|
|
240
|
+
| `compression` | `str \| None` | `gzip`、`deflate` 或 `none` |
|
|
241
|
+
| `headers` | `dict \| None` | OTLP 导出器请求头 |
|
|
242
|
+
| `extra_resource_attributes` | `dict \| None` | 额外的资源属性 |
|
|
243
|
+
| `span_processors` | `list \| None` | 自定义 SpanProcessor 列表 |
|
|
244
|
+
| `sampler` | `Sampler \| None` | 自定义采样器实例 |
|
|
245
|
+
| `disabled_instrumentations` | `list \| None` | 禁用的 instrumentor 名称列表,`"*"` 禁用全部 |
|
|
246
|
+
|
|
247
|
+
**返回值:** `list[str]` — 成功加载的 instrumentor 名称列表。
|
|
248
|
+
|
|
249
|
+
### `fastx_opentelemetry.list_available_instrumentors()`
|
|
250
|
+
|
|
251
|
+
返回所有已安装的 instrumentor 名称列表。
|
|
252
|
+
|
|
253
|
+
### `fastx_opentelemetry.suggest_missing_instrumentations(loaded)`
|
|
254
|
+
|
|
255
|
+
检查已安装但未增强的库,返回建议安装的 instrumentation 包名列表。
|
|
256
|
+
|
|
257
|
+
### 采样器
|
|
258
|
+
|
|
259
|
+
#### `RulesBasedSampler(rules, default_sampler=None)`
|
|
260
|
+
|
|
261
|
+
按规则匹配 span 名称或属性,委托给不同的采样器。
|
|
262
|
+
|
|
263
|
+
#### `SamplingRule(sampler, match_name=None, match_attributes=None)`
|
|
264
|
+
|
|
265
|
+
单条采样规则。`match_name` 为正则表达式匹配 span 名称,`match_attributes` 为属性键值对匹配。
|
|
266
|
+
|
|
267
|
+
#### `RateLimitingSampler(max_spans_per_second)`
|
|
268
|
+
|
|
269
|
+
基于令牌桶算法的限流采样器,限制每秒采样的 span 数量。
|
|
270
|
+
|
|
271
|
+
### Baggage 传播与 Span 处理
|
|
272
|
+
|
|
273
|
+
#### `FastxBaggagePropagator`
|
|
274
|
+
|
|
275
|
+
包装 W3C Baggage 传播器,在 `extract` 时自动转换服务调用链。`init()` 时自动设置为全局传播器。
|
|
276
|
+
|
|
277
|
+
#### `FastxSpanProcessor`
|
|
278
|
+
|
|
279
|
+
在 Span `on_start` 时将所有 Baggage 条目复制为 Span 属性。`init()` 时自动添加到 TracerProvider。
|
|
280
|
+
|
|
281
|
+
## 本地开发
|
|
282
|
+
|
|
283
|
+
```bash
|
|
284
|
+
# 安装依赖
|
|
285
|
+
pip install -e ".[all]"
|
|
286
|
+
|
|
287
|
+
# 运行示例
|
|
288
|
+
export OTEL_LOG_LEVEL=debug
|
|
289
|
+
export OTEL_EXPORTER_OTLP_ENDPOINT=http://127.0.0.1:4318
|
|
290
|
+
export OTEL_SERVICE_NAME=my-app
|
|
291
|
+
|
|
292
|
+
python examples/app.py
|
|
293
|
+
```
|
|
294
|
+
|
|
295
|
+
## License
|
|
296
|
+
|
|
297
|
+
Apache-2.0
|