register a dataset with the tf.data service, and another process to consume
    data from the dataset.
*   Adds support for dispatcher fault tolerance. To enable fault tolerance,
    configure a `work_dir` when running your dispatcher server and set
    `dispatcher_fault_tolerance=True`. The dispatcher will store its state to

            assertNotNull(result);
            assertEquals(1, result.intValue());
        } catch (Exception e) {
            // Expected behavior when components are not fully initialized
            assertTrue(e.getCause() instanceof NullPointerException || e.getCause() instanceof ContainerNotAvailableException);
        } finally {
            tempPropFile.delete();
        }
    }

  // To test this, we give an input where bit 7 is not cleared. For example:
  // (1) 00 01 02 03 04 05 06 07 80
  // (2) 00 01 02 03 04 05 06 07 81
  // (3) 00 01 02 03 04 05 06 07 ff (or anything in between)
  // A fault implementation will generate collisions for these inputs.
  public void testCollisionsDueToIncorrectSignExtension() {
    byte[] col1 = new byte[] {0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, (byte) 0x80};

  // To test this, we give an input where bit 7 is not cleared. For example:
  // (1) 00 01 02 03 04 05 06 07 80
  // (2) 00 01 02 03 04 05 06 07 81
  // (3) 00 01 02 03 04 05 06 07 ff (or anything in between)
  // A fault implementation will generate collisions for these inputs.
  public void testCollisionsDueToIncorrectSignExtension() {
    byte[] col1 = new byte[] {0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, (byte) 0x80};

**Resource Management**: `AutoCloseable` throughout, `DeferredFileOutputStream` for large responses, connection pooling, background temp file deletion via `FileUtil.deleteInBackground()`

**Fault Tolerance**: `FaultTolerantClient` wrapper (retry, circuit breaker), `SwitchableHttpClient` for HTTP client fallback

---

## Key Components

### Core Classes

        for sub_annotation in get_args(annotation)
    )


def is_pydantic_v1_model_instance(obj: Any) -> bool:
    # TODO: remove this function once the required version of Pydantic fully
    # removes pydantic.v1
    try:
        with warnings.catch_warnings():
            warnings.simplefilter("ignore", UserWarning)
            from pydantic import v1
    except ImportError:  # pragma: no cover

        urlIterator.remove()
      }
    }
```

### Troubleshooting

1. Valid cacheable responses are not being cached

Make sure you are reading responses fully as unless they are read fully, cancelled or stalled Responses will not be cached.

### Overriding normal cache behaviour

    * Si on compare Uvicorn, il faut le comparer à d'autres serveurs d'applications comme Daphne, Hypercorn, uWSGI, etc.
* **Starlette** :

Embedding a Java runtime in the distribution would provide some benefits, such as allowing the Launcher, Daemon and Workers to run on it, removing the prerequisite of an installed Java runtime.
However, this does not fully remove the prerequisite, as the Wrapper itself would still need an installed Java runtime to execute.

Gradle needs a solution that covers the use-case of starting the Wrapper itself as well given its importance.

Por exemplo, se o cliente tentar enviar um cookie `santa_tracker` com o valor de `good-list-please`, o cliente receberá uma resposta de **erro** informando que o `santa_tracker` <dfn title="O papai noel desaprova a falta de biscoitos. 🎅 Ok, chega de piadas com os cookies.">cookie não é permitido</dfn>:

```json
{
    "detail": [
        {
            "type": "extra_forbidden",
            "loc": ["cookie", "santa_tracker"],