import org.opensearch.monitor.jvm.JvmStats;
import org.opensearch.monitor.jvm.JvmStats.BufferPool;
import org.opensearch.monitor.jvm.JvmStats.Classes;
import org.opensearch.monitor.jvm.JvmStats.GarbageCollectors;
import org.opensearch.monitor.jvm.JvmStats.Mem;
import org.opensearch.monitor.jvm.JvmStats.Threads;
import org.opensearch.monitor.os.OsProbe;
import org.opensearch.monitor.os.OsStats;

Para isso, você pode utilizar o `WSGIMiddleware` para encapsular a sua aplicação WSGI, como por exemplo Flask, Django, etc.

## Usando o `WSGIMiddleware`

Você precisa importar o `WSGIMiddleware`.

Em seguinda, encapsular a aplicação WSGI (e.g. Flask) com o middleware.

E então **"montar"** em um caminho de rota.

```Python hl_lines="2-3  23"

	m        *Monitor
	opts     *MonitorReaderOptions
}

// BucketOptions represents the bucket and optionally its replication target pair.
type BucketOptions struct {
	Name           string
	ReplicationARN string // This is optional, and not mandatory.
}

// MonitorReaderOptions provides configurable options for monitor reader implementation.
type MonitorReaderOptions struct {

    async with AsyncClient(
        transport=ASGITransport(app=app), base_url="http://test"
    ) as ac:
        response = await ac.get("/")
    assert response.status_code == 200

import org.lastaflute.di.core.external.GenericExternalContextComponentDefRegister;
import org.lastaflute.di.core.factory.SingletonLaContainerFactory;
import org.opensearch.monitor.jvm.JvmInfo;
import org.opensearch.monitor.os.OsProbe;
import org.opensearch.monitor.process.ProcessProbe;

import jakarta.annotation.Resource;

public class ThumbnailGenerator {

Se você precisar ter duas aplicações FastAPI independentes, cada uma com seu próprio OpenAPI e suas próprias interfaces de documentação, você pode ter um aplicativo principal e "montar" uma (ou mais) sub-aplicações.

## Montando uma aplicação **FastAPI**

    /* the implementation class to collect information from task execution result */
    klass: Class<T>,
    /* which tasks we need to monitor? For example, cache-miss-monitor monitors `AbstractCompile` tasks */
    taskFilter: (Task) -> Boolean,
    /* pass the collected information in build-logic and main build to build scan */

`<Address for MinIO Node>/minio/v2/metrics/resource`.

To use this endpoint, setup Prometheus to scrape data from this endpoint. Read more on how to configure and use Prometheus to monitor MinIO server in [How to monitor MinIO server with Prometheus](https://github.com/minio/minio/blob/master/docs/metrics/prometheus/README.md).

### **Deprecated metrics monitoring**

Der beste Ansatz bestand außerdem darin, bereits bestehende Standards zu nutzen.

Bevor ich also überhaupt angefangen habe, **FastAPI** zu schreiben, habe ich mehrere Monate damit verbracht, die Spezifikationen für OpenAPI, JSON Schema, OAuth2, usw. zu studieren und deren Beziehungen, Überschneidungen und Unterschiede zu verstehen.

## Design

from fastapi import FastAPI

app = FastAPI()


@app.get("/")
async def root():