util.concurrent tests under Guava.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see
    https://groups.google.com/g/google-web-toolkit-contributors/c/CqYH59Dt_rQ/m/uVGW1QdUsXUJ
    for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one

    util.concurrent tests under Guava.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see
    https://groups.google.com/g/google-web-toolkit-contributors/c/CqYH59Dt_rQ/m/uVGW1QdUsXUJ
    for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one

### 並行 + 平行: Web + 機器學習 { #concurrency-parallelism-web-machine-learning }

使用 **FastAPI**，你可以利用並行的優勢，這在 Web 開發中非常常見（這也是 NodeJS 的最大吸引力）。

但你也可以利用平行與多行程 (multiprocessing)（讓多個行程同時運行） 的優勢來處理機器學習系統中的 **CPU 密集型**工作。

這一點，再加上 Python 是 **資料科學**、機器學習，尤其是深度學習的主要語言，讓 **FastAPI** 成為資料科學/機器學習 Web API 和應用程式（以及許多其他應用程式）的絕佳選擇。

想了解如何在生產環境中實現這種平行性，請參見 [部屬](deployment/index.md)。

### 并发 + 并行: Web + 机器学习 { #concurrency-parallelism-web-machine-learning }

使用 **FastAPI**，你可以利用 Web 开发中常见的并发机制的优势（NodeJS 的主要吸引力）。

并且，你也可以利用并行和多进程（让多个进程并行运行）的优点来处理与机器学习系统中类似的 **CPU 密集型** 工作。

这一点，再加上 Python 是**数据科学**、机器学习（尤其是深度学习）的主要语言这一简单事实，使得 **FastAPI** 与数据科学/机器学习 Web API 和应用程序（以及其他许多应用程序）非常匹配。

了解如何在生产环境中实现这种并行性，可查看此文 [部署](deployment/index.md)。

    util.concurrent tests under Guava.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see
    https://groups.google.com/g/google-web-toolkit-contributors/c/CqYH59Dt_rQ/m/uVGW1QdUsXUJ
    for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one

```yaml
route:
  group_by: ['alertname']
  group_wait: 30s
  group_interval: 5m
  repeat_interval: 1h
  receiver: 'web.hook'
receivers:
  - name: 'web.hook'
    webhook_configs:
      - url: 'http://127.0.0.1:8010/webhook'
inhibit_rules:
  - source_match:
      severity: 'critical'
    target_match:
      severity: 'warning'

    util.concurrent tests under Guava.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see
    https://groups.google.com/g/google-web-toolkit-contributors/c/CqYH59Dt_rQ/m/uVGW1QdUsXUJ
    for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one

    util.concurrent tests under Guava.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see
    https://groups.google.com/g/google-web-toolkit-contributors/c/CqYH59Dt_rQ/m/uVGW1QdUsXUJ
    for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one

    util.concurrent tests under Guava.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see
    https://groups.google.com/g/google-web-toolkit-contributors/c/CqYH59Dt_rQ/m/uVGW1QdUsXUJ
    for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one

    util.concurrent tests under Guava.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see
    https://groups.google.com/g/google-web-toolkit-contributors/c/CqYH59Dt_rQ/m/uVGW1QdUsXUJ
    for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one