<span style="background-color:#C4A000"><font color="#2E3436">│                                                     │</font></span>
 <span style="background-color:#C4A000"><font color="#2E3436">│  Running in development mode, for production use:   │</font></span>
 <span style="background-color:#C4A000"><font color="#2E3436">│                                                     │</font></span>

 L'option `--reload` consomme beaucoup plus de ressources, est plus instable, etc.

 Cela aide beaucoup pendant le **développement**, mais vous **ne devriez pas** l'utiliser en **production**.

///

## Hypercorn avec Trio

Starlette et **FastAPI** sont basés sur
<a href="https://anyio.readthedocs.io/en/stable/" class="external-link" target="_blank">AnyIO</a>, qui les rend

But you can also provide other alternative `servers`, for example if you want *the same* docs UI to interact with both a staging and a production environment.

If you pass a custom list of `servers` and there's a `root_path` (because your API lives behind a proxy), **FastAPI** will insert a "server" with this `root_path` at the beginning of the list.

For example:

* SQLite
* Oracle
* Microsoft SQL Server, etc.

In this example, we'll use **SQLite**, because it uses a single file and Python has integrated support. So, you can copy this example and run it as is.

Later, for your production application, you might want to use a database server like **PostgreSQL**.

/// tip

import jcifs.CIFSException;

/**
 * Default implementation of SMB3 compression service.
 *
 * Provides compression using Java's built-in deflate algorithm as a base
 * for LZ77-based compression. For production use, more specialized
 * implementations of LZNT1 and Pattern_V1 algorithms would be needed.
 */
public class DefaultCompressionService implements CompressionService {

INHERIT: ../en/mkdocs.maybe-insiders.yml
site_name: FastAPI
site_description: FastAPI framework, high performance, easy to learn, fast to code, ready for production
site_url: https://fastapi.tiangolo.com/
theme:
  name: material
  custom_dir: ../en/overrides
  palette:
  - media: (prefers-color-scheme)
    toggle:
      icon: material/lightbulb-auto
      name: Switch to light mode
  - media: '(prefers-color-scheme: light)'
    scheme: default

```console
$ <font color="#4E9A06">fastapi</font> run --workers 4 <u style="text-decoration-style:solid">main.py</u>

  <span style="background-color:#009485"><font color="#D3D7CF"> FastAPI </font></span>  Starting production server 🚀

             Searching for package file structure from directories with
             <font color="#3465A4">__init__.py</font> files

 <span style="background-color:#C4A000"><font color="#2E3436">│                                                     │</font></span>
 <span style="background-color:#C4A000"><font color="#2E3436">│  Running in development mode, for production use:   │</font></span>
 <span style="background-color:#C4A000"><font color="#2E3436">│                                                     │</font></span>

        {
            "url": "https://stag.example.com",
            "description": "Staging environment"
        },
        {
            "url": "https://prod.example.com",
            "description": "Production environment"
        }
    ],
    "paths": {
            // More stuff here
    }
}
```

/// tip

👀 🚘-🏗 💽 ⏮️ `url` 💲 `/api/v1`, ✊ ⚪️➡️ `root_path`.

///

[**sandbox**](https://developers.google.com/code-sandboxing). Memory corruptions
in TensorFlow ops can be recognized as security issues only if they are
reachable and exploitable through production-grade, benign models.

### Compilation

Compiling models via the recommended entry points described in
[XLA](https://www.tensorflow.org/xla) and
[JAX](https://jax.readthedocs.io/en/latest/jax-101/02-jitting.html)