azure-monitor-ingestion-py
用 Python 透過 Logs Ingestion API 把自訂 log 送進 Log Analytics
中文翻譯
---
name: azure-monitor-ingestion-py
description: |
Python 版 Azure Monitor Ingestion SDK。用於透過 Logs Ingestion API 把自訂 log 送到 Log Analytics workspace。
Triggers: "azure-monitor-ingestion", "LogsIngestionClient", "custom logs", "DCR", "data collection rule", "Log Analytics".
license: MIT
metadata:
author: Microsoft
version: "1.0.0"
package: azure-monitor-ingestion
---
# Azure Monitor Ingestion SDK for Python
用 Logs Ingestion API 把自訂 log 送到 Azure Monitor Log Analytics workspace。
## 安裝
```bash
pip install azure-monitor-ingestion
pip install azure-identity
```
## 環境變數
```bash
# Data Collection Endpoint (DCE)
AZURE_DCE_ENDPOINT=https://<dce-name>.<region>.ingest.monitor.azure.com # 所有驗證方式都必填
# Data Collection Rule (DCR) 的 immutable ID
AZURE_DCR_RULE_ID=dcr-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # 所有驗證方式都必填
# DCR 的 stream 名稱
AZURE_DCR_STREAM_NAME=Custom-MyTable_CL # 所有驗證方式都必填
AZURE_TOKEN_CREDENTIALS=prod # 只有在正式環境使用 DefaultAzureCredential 時才需要
```
## 前置條件
使用這個 SDK 之前，你需要：
1. **Log Analytics Workspace**：log 的目的地
2. **Data Collection Endpoint (DCE)**：ingestion endpoint
3. **Data Collection Rule (DCR)**：定義 schema 與目的地
4. **自訂資料表**：在 Log Analytics 裡（透過 DCR 或手動建立）
## 驗證與生命週期
> **🔑 下面每一段程式碼都適用兩條規則：**
>
> 1. **優先使用 `DefaultAzureCredential`。** 它在本機（Azure CLI / VS Code / Developer CLI）和 Azure 上（managed identity、workload identity）都能運作，不用改程式碼。避免使用 connection string、account/API key，它們會繞過 Entra 的稽核與輪替。
> - 本機開發：直接用 `DefaultAzureCredential` 即可。
> - 正式環境：設定 `AZURE_TOKEN_CREDENTIALS=prod`（或 `AZURE_TOKEN_CREDENTIALS=<specific_credential>`），把 credential chain 限制在適合正式環境的 credential。
> 2. **每個 client 都要包在 context manager 裡**，這樣 HTTP transport、socket 和 token cache 才會確定性地被釋放：
> - 同步：`with <Client>(...) as client:`
> - 非同步：`async with <Client>(...) as client:` **以及** `async with DefaultAzureCredential() as credential:`（來自 `azure.identity.aio`）
>
> 範例程式碼可能會省略這段設定，但正式環境的程式碼一定要同時遵守這兩條規則。
```python
from azure.monitor.ingestion import LogsIngestionClient
from azure.identity import DefaultAzureCredential, ManagedIdentityCredential
import os
# 本機開發：DefaultAzureCredential。正式環境：設定 AZURE_TOKEN_CREDENTIALS=prod 或 AZURE_TOKEN_CREDENTIALS=<specific_credential>
credential = DefaultAzureCredential(require_envvar=True)
# 或在正式環境直接使用特定的 credential：
# See https://learn.microsoft.com/python/api/overview/azure/identity-readme?view=azure-python#credential-classes
# credential = ManagedIdentityCredential()
with LogsIngestionClient(
endpoint=os.environ["AZURE_DCE_ENDPOINT"],
credential=credential
) as client:
# 後續所有操作都用 `client.upload(...)`（見下方範例）
...
```
## 上傳自訂 Log
```python
from azure.monitor.ingestion import LogsIngestionClient
from azure.identity import DefaultAzureCredential
import os
rule_id = os.environ["AZURE_DCR_RULE_ID"]
stream_name = os.environ["AZURE_DCR_STREAM_NAME"]
logs = [
{"TimeGenerated": "2024-01-15T10:00:00Z", "Computer": "server1", "Message": "Application started"},
{"TimeGenerated": "2024-01-15T10:01:00Z", "Computer": "server1", "Message": "Processing request"},
{"TimeGenerated": "2024-01-15T10:02:00Z", "Computer": "server2", "Message": "Connection established"}
]
with LogsIngestionClient(
endpoint=os.environ["AZURE_DCE_ENDPOINT"],
credential=DefaultAzureCredential()
) as client:
client.upload(rule_id=rule_id, stream_name=stream_name, logs=logs)
```
## 從 JSON 檔上傳
```python
import json
with open("logs.json", "r") as f:
logs = json.load(f)
client.upload(rule_id=rule_id, stream_name=stream_name, logs=logs)
```
## 自訂錯誤處理
用 callback 處理部分失敗：
```python
failed_logs = []
def on_error(error):
print(f"Upload failed: {error.error}")
failed_logs.extend(error.failed_logs)
client.upload(
rule_id=rule_id,
stream_name=stream_name,
logs=logs,
on_error=on_error
)
# 重試失敗的 log
if failed_logs:
print(f"Retrying {len(failed_logs)} failed logs...")
client.upload(rule_id=rule_id, stream_name=stream_name, logs=failed_logs)
```
## 忽略錯誤
```python
def ignore_errors(error):
pass # 靜默忽略上傳失敗
client.upload(
rule_id=rule_id,
stream_name=stream_name,
logs=logs,
on_error=ignore_errors
)
```
## 非同步 Client
```python
import asyncio
from azure.monitor.ingestion.aio import LogsIngestionClient
from azure.identity.aio import DefaultAzureCredential
async def upload_logs():
async with LogsIngestionClient(
endpoint=endpoint,
credential=DefaultAzureCredential()
) as client:
await client.upload(
rule_id=rule_id,
stream_name=stream_name,
logs=logs
)
asyncio.run(upload_logs())
```
## 主權雲（Sovereign Clouds）
```python
from azure.identity import AzureAuthorityHosts, DefaultAzureCredential
from azure.monitor.ingestion import LogsIngestionClient
# Azure Government
credential = DefaultAzureCredential(authority=AzureAuthorityHosts.AZURE_GOVERNMENT)
with LogsIngestionClient(
endpoint="https://example.ingest.monitor.azure.us",
credential=credential,
credential_scopes=["https://monitor.azure.us/.default"]
) as client:
# client.upload(...)
...
```
## 批次行為
SDK 會自動：
- 把 log 切成 1MB 以下的區塊
- 用 gzip 壓縮每個區塊
- 平行上傳區塊
大量 log 不需要自己做批次。
## Client 類型
| Client | 用途 |
|--------|---------|
| `LogsIngestionClient` | 上傳 log 的同步 client |
| `LogsIngestionClient` (aio) | 上傳 log 的非同步 client |
## 關鍵概念
| 概念 | 說明 |
|---------|-------------|
| **DCE** | Data Collection Endpoint，ingestion URL |
| **DCR** | Data Collection Rule，定義 schema、轉換與目的地 |
| **Stream** | DCR 內具名的資料流 |
| **Custom Table** | Log Analytics 裡的目標資料表（名稱以 `_CL` 結尾） |
## DCR Stream 名稱格式
Stream 名稱遵循以下模式：
- `Custom-<TableName>_CL`：用於自訂資料表
- `Microsoft-<TableName>`：用於內建資料表
## 最佳實務
1. **同步或非同步擇一，並保持一致。** 在同一條呼叫路徑裡，不要混用 `azure.xxx` 同步 client 和 `azure.xxx.aio` 非同步 client。每個 module 選一種模式。
2. **client 和非同步 credential 一律用 context manager。** 每個 client 都包在 `with Client(...) as client:`（同步）或 `async with Client(...) as client:`（非同步）裡，確保正確清理。非同步的 `DefaultAzureCredential`（來自 `azure.identity.aio`）也要用 `async with credential:`，這樣 token 和 transport 才會被清掉。
3. **在本機執行的程式碼用 `DefaultAzureCredential`**。在 Azure 上執行的程式碼用特定的 token credential。
4. **優雅地處理錯誤**：部分失敗用 `on_error` callback
5. **包含 TimeGenerated**：所有 log 都必填的欄位
6. **符合 DCR schema**：log 欄位要符合 DCR 的欄位定義
7. **高吞吐量情境用非同步 client**
8. **批次上傳**：SDK 會處理批次，但送出的區塊大小要合理
9. **監控 ingestion**：到 Log Analytics 檢查 ingestion 狀態
## 參考檔案
| 檔案 | 內容 |
|------|----------|
| [references/capabilities.md](references/capabilities.md) | 其他非主打功能、operation group 涵蓋範圍，以及正式環境檢查清單。 |
| [references/non-hero-scenarios.md](references/non-hero-scenarios.md) | 針對次要／進階情境的專屬非主打範例。 |英文原文
---
name: azure-monitor-ingestion-py
description: |
Azure Monitor Ingestion SDK for Python. Use for sending custom logs to Log Analytics workspace via Logs Ingestion API.
Triggers: "azure-monitor-ingestion", "LogsIngestionClient", "custom logs", "DCR", "data collection rule", "Log Analytics".
license: MIT
metadata:
author: Microsoft
version: "1.0.0"
package: azure-monitor-ingestion
---
# Azure Monitor Ingestion SDK for Python
Send custom logs to Azure Monitor Log Analytics workspace using the Logs Ingestion API.
## Installation
```bash
pip install azure-monitor-ingestion
pip install azure-identity
```
## Environment Variables
```bash
# Data Collection Endpoint (DCE)
AZURE_DCE_ENDPOINT=https://<dce-name>.<region>.ingest.monitor.azure.com # Required for all auth methods
# Data Collection Rule (DCR) immutable ID
AZURE_DCR_RULE_ID=dcr-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # Required for all auth methods
# Stream name from DCR
AZURE_DCR_STREAM_NAME=Custom-MyTable_CL # Required for all auth methods
AZURE_TOKEN_CREDENTIALS=prod # Required only if DefaultAzureCredential is used in production
```
## Prerequisites
Before using this SDK, you need:
1. **Log Analytics Workspace** — Target for your logs
2. **Data Collection Endpoint (DCE)** — Ingestion endpoint
3. **Data Collection Rule (DCR)** — Defines schema and destination
4. **Custom Table** — In Log Analytics (created via DCR or manually)
## Authentication & Lifecycle
> **🔑 Two rules apply to every code sample below:**
>
> 1. **Prefer `DefaultAzureCredential`.** It works locally (Azure CLI / VS Code / Developer CLI) and in Azure (managed identity, workload identity) with no code change. Avoid connection strings, account/API keys — they bypass Entra audit and rotation.
> - Local dev: `DefaultAzureCredential` works as-is.
> - Production: set `AZURE_TOKEN_CREDENTIALS=prod` (or `AZURE_TOKEN_CREDENTIALS=<specific_credential>`) to constrain the credential chain to production-safe credentials.
> 2. **Wrap every client in a context manager** so HTTP transports, sockets, and token caches are released deterministically:
> - Sync: `with <Client>(...) as client:`
> - Async: `async with <Client>(...) as client:` **and** `async with DefaultAzureCredential() as credential:` (from `azure.identity.aio`)
>
> Snippets may abbreviate this setup, but production code should always follow both rules.
```python
from azure.monitor.ingestion import LogsIngestionClient
from azure.identity import DefaultAzureCredential, ManagedIdentityCredential
import os
# Local dev: DefaultAzureCredential. Production: set AZURE_TOKEN_CREDENTIALS=prod or AZURE_TOKEN_CREDENTIALS=<specific_credential>
credential = DefaultAzureCredential(require_envvar=True)
# Or use a specific credential directly in production:
# See https://learn.microsoft.com/python/api/overview/azure/identity-readme?view=azure-python#credential-classes
# credential = ManagedIdentityCredential()
with LogsIngestionClient(
endpoint=os.environ["AZURE_DCE_ENDPOINT"],
credential=credential
) as client:
# Use `client.upload(...)` for all subsequent operations (see examples below)
...
```
## Upload Custom Logs
```python
from azure.monitor.ingestion import LogsIngestionClient
from azure.identity import DefaultAzureCredential
import os
rule_id = os.environ["AZURE_DCR_RULE_ID"]
stream_name = os.environ["AZURE_DCR_STREAM_NAME"]
logs = [
{"TimeGenerated": "2024-01-15T10:00:00Z", "Computer": "server1", "Message": "Application started"},
{"TimeGenerated": "2024-01-15T10:01:00Z", "Computer": "server1", "Message": "Processing request"},
{"TimeGenerated": "2024-01-15T10:02:00Z", "Computer": "server2", "Message": "Connection established"}
]
with LogsIngestionClient(
endpoint=os.environ["AZURE_DCE_ENDPOINT"],
credential=DefaultAzureCredential()
) as client:
client.upload(rule_id=rule_id, stream_name=stream_name, logs=logs)
```
## Upload from JSON File
```python
import json
with open("logs.json", "r") as f:
logs = json.load(f)
client.upload(rule_id=rule_id, stream_name=stream_name, logs=logs)
```
## Custom Error Handling
Handle partial failures with a callback:
```python
failed_logs = []
def on_error(error):
print(f"Upload failed: {error.error}")
failed_logs.extend(error.failed_logs)
client.upload(
rule_id=rule_id,
stream_name=stream_name,
logs=logs,
on_error=on_error
)
# Retry failed logs
if failed_logs:
print(f"Retrying {len(failed_logs)} failed logs...")
client.upload(rule_id=rule_id, stream_name=stream_name, logs=failed_logs)
```
## Ignore Errors
```python
def ignore_errors(error):
pass # Silently ignore upload failures
client.upload(
rule_id=rule_id,
stream_name=stream_name,
logs=logs,
on_error=ignore_errors
)
```
## Async Client
```python
import asyncio
from azure.monitor.ingestion.aio import LogsIngestionClient
from azure.identity.aio import DefaultAzureCredential
async def upload_logs():
async with LogsIngestionClient(
endpoint=endpoint,
credential=DefaultAzureCredential()
) as client:
await client.upload(
rule_id=rule_id,
stream_name=stream_name,
logs=logs
)
asyncio.run(upload_logs())
```
## Sovereign Clouds
```python
from azure.identity import AzureAuthorityHosts, DefaultAzureCredential
from azure.monitor.ingestion import LogsIngestionClient
# Azure Government
credential = DefaultAzureCredential(authority=AzureAuthorityHosts.AZURE_GOVERNMENT)
with LogsIngestionClient(
endpoint="https://example.ingest.monitor.azure.us",
credential=credential,
credential_scopes=["https://monitor.azure.us/.default"]
) as client:
# client.upload(...)
...
```
## Batching Behavior
The SDK automatically:
- Splits logs into chunks of 1MB or less
- Compresses each chunk with gzip
- Uploads chunks in parallel
No manual batching needed for large log sets.
## Client Types
| Client | Purpose |
|--------|---------|
| `LogsIngestionClient` | Sync client for uploading logs |
| `LogsIngestionClient` (aio) | Async client for uploading logs |
## Key Concepts
| Concept | Description |
|---------|-------------|
| **DCE** | Data Collection Endpoint — ingestion URL |
| **DCR** | Data Collection Rule — defines schema, transformations, destination |
| **Stream** | Named data flow within a DCR |
| **Custom Table** | Target table in Log Analytics (ends with `_CL`) |
## DCR Stream Name Format
Stream names follow patterns:
- `Custom-<TableName>_CL` — For custom tables
- `Microsoft-<TableName>` — For built-in tables
## Best Practices
1. **Pick sync OR async and stay consistent.** Do not mix `azure.xxx` sync clients with `azure.xxx.aio` async clients in the same call path. Choose one mode per module.
2. **Always use context managers for clients and async credentials.** Wrap every client in `with Client(...) as client:` (sync) or `async with Client(...) as client:` (async) to ensure proper cleanup. For async `DefaultAzureCredential` from `azure.identity.aio`, also use `async with credential:` so tokens and transports are cleaned up.
3. **Use `DefaultAzureCredential`** for code that runs locally. Use a specific token credential for code that runs in Azure.
4. **Handle errors gracefully** — use `on_error` callback for partial failures
5. **Include TimeGenerated** — Required field for all logs
6. **Match DCR schema** — Log fields must match DCR column definitions
7. **Use async client** for high-throughput scenarios
8. **Batch uploads** — SDK handles batching, but send reasonable chunks
9. **Monitor ingestion** — Check Log Analytics for ingestion status
## Reference Files
| File | Contents |
|------|----------|
| [references/capabilities.md](references/capabilities.md) | Additional non-hero capabilities, operation-group coverage, and production checklists. |
| [references/non-hero-scenarios.md](references/non-hero-scenarios.md) | Dedicated non-hero examples for secondary/advanced scenarios. |