When referring to the remote machine, it's common to call it **server**, but also **machine**, **VM** (virtual machine), **node**. Those all refer to some type of remote machine, normally running Linux, where you run programs.

## Install the Server Program

When you install FastAPI, it comes with a production server, Uvicorn, and you can start it with the `fastapi run` command.

But you can also install an ASGI server manually:

=== "Uvicorn"

### Server Memory

For example, if your code loads a Machine Learning model with **1 GB in size**, when you run one process with your API, it will consume at least 1 GB of RAM. And if you start **4 processes** (4 workers), each will consume 1 GB of RAM. So in total, your API will consume **4 GB of RAM**.

And if your remote server or virtual machine only has 3 GB of RAM, trying to load more than 4 GB of RAM will cause problems. 🚨

Linux containers run using the same Linux kernel of the host (machine, virtual machine, cloud server, etc). This just means that they are very lightweight (compared to full virtual machines emulating an entire operating system).

This way, containers consume **little resources**, an amount comparable to running the processes directly (a virtual machine would consume much more).

By default it will listen on the IP address `127.0.0.1`, which is the IP for your machine to communicate with itself alone (`localhost`).

## `fastapi run`

When you run `fastapi run`, it will run on production mode by default.

It will have **auto-reload disabled** by default.

> **NOTE:** Do **NOT** run `gradle build` on the local development environment,
> even if you have Gradle or Develocity build caching enabled for the project.
> The Gradle Build Tool repository is massive, and it will take ages to build on
> a local machine without necessary parallelization and caching.
> The full test suites are executed on the CI instance for multiple configurations,

build:rbe_cross_compile_base --remote_instance_name=projects/tensorflow-testing/instances/default_instance

# XLA depends on some local Python headers that are configured as Genrule. They
# are present on the local host machine but not on the remote execution machine,
# leading to build failures. To resolve the issue, the following line is added
# to make sure all Genrule targets are excuted locally.
build:rbe_cross_compile_base_xla --config=rbe_cross_compile_base

server["Server on http://127.0.0.1:8000/app"]

browser --> proxy
proxy --> server
```

!!! tip
    The IP `0.0.0.0` is commonly used to mean that the program listens on all the IPs available in that machine/server.

The docs UI would also need the OpenAPI schema to declare that this API `server` is located at `/api/v1` (behind the proxy). For example:

```JSON hl_lines="4-8"
{
    "openapi": "3.1.0",

In the output, there's a line with something like:

```hl_lines="4"
INFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
```

That line shows the URL where your app is being served, in your local machine.

### Check it

Open your browser at <a href="http://127.0.0.1:8000" class="external-link" target="_blank">http://127.0.0.1:8000</a>.

You will see the JSON response as:

```JSON
{"message": "Hello World"}

## Görüşler

"_[...] Bugünlerde **FastAPI**'ı çok fazla kullanıyorum. [...] Aslında bunu ekibimin **Microsoft'taki Machine Learning servislerinin** tamamında kullanmayı planlıyorum. Bunlardan bazıları **Windows**'un ana ürünlerine ve **Office** ürünlerine entegre ediliyor._"

## Opiniões

"*[...] Estou usando **FastAPI** muito esses dias. [...] Estou na verdade planejando utilizar ele em todos os times de **serviços _Machine Learning_ na Microsoft**. Alguns deles estão sendo integrados no _core_ do produto **Windows** e alguns produtos **Office**.*"