/// info

If you are using containers, for example with Docker or Kubernetes, I'll tell you more about that in the next chapter: [FastAPI in Containers - Docker](docker.md){.internal-link target=_blank}.

In particular, when running on **Kubernetes** you will probably **not** want to use workers and instead run **a single Uvicorn process per container**, but I'll tell you about it later in that chapter.

///

## Multiple Workers { #multiple-workers }

## In a hurry? { #in-a-hurry }

<abbr title="too long; didn't read"><strong>TL;DR:</strong></abbr>

If you are using third party libraries that tell you to call them with `await`, like:

```Python
results = await some_library()
```

Then, declare your *path operation functions* with `async def` like:

```Python hl_lines="2"
@app.get('/')

If you want to secure your API, there are several better things you can do, for example:

* Make sure you have well defined Pydantic models for your request bodies and responses.
* Configure any required permissions and roles using dependencies.
* Never store plaintext passwords, only password hashes.
* Implement and use well-known cryptographic tools, like Passlib and JWT tokens, etc.

In general, ASGI middlewares are classes that expect to receive an ASGI app as the first argument.

So, in the documentation for third-party ASGI middlewares they will probably tell you to do something like:

```Python
from unicorn import UnicornMiddleware

app = SomeASGIApp()

new_app = UnicornMiddleware(app, some_config="rainbow")
```

I'll tell you a bit more about these **concepts** here, and that would hopefully give you the **intuition** you would need to decide how to deploy your API in very different environments, possibly even in **future** ones that don't exist yet.

    *   tf.optimizers => tf.keras.optimizers
    *   Add tf.keras.layers.AbstractRNNCell as the preferred implementation of
        RNN cell for TF v2. User can use it to implement RNN cell with custom
        behavior.
    *   Adding `clear_losses` API to be able to clear losses at the end of
        forward pass in a custom training loop in eager.

Keep in mind that this means that the whole contents will be stored in memory. This will work well for small files.

But there are several cases in which you might benefit from using `UploadFile`.

## File Parameters with `UploadFile` { #file-parameters-with-uploadfile }

## Declare `Cookie` parameters { #declare-cookie-parameters }

Then declare the cookie parameters using the same structure as with `Path` and `Query`.

You can define the default value as well as all the extra validation or annotation parameters:

{* ../../docs_src/cookie_params/tutorial001_an_py310.py hl[9] *}

/// note | Technical Details

	opts.ReplicateILMExpiry = r.Form.Get("replicateILMExpiry") == "true"
	return
}

// SRPeerJoin - PUT /minio/admin/v3/site-replication/join
//
// used internally to tell current cluster to enable SR with
// the provided peer clusters and service account.
func (a adminAPIHandlers) SRPeerJoin(w http.ResponseWriter, r *http.Request) {
	ctx := r.Context()

Actually, `Query`, `Path` and others you'll see next create objects of subclasses of a common `Param` class, which is itself a subclass of Pydantic's `FieldInfo` class.

And Pydantic's `Field` returns an instance of `FieldInfo` as well.

`Body` also returns objects of a subclass of `FieldInfo` directly. And there are others you will see later that are subclasses of the `Body` class.