但你也可以利用平行與多行程 (multiprocessing)（讓多個行程同時運行） 的優勢來處理機器學習系統中的 **CPU 密集型**工作。

這一點，再加上 Python 是 **資料科學**、機器學習，尤其是深度學習的主要語言，讓 **FastAPI** 成為資料科學/機器學習 Web API 和應用程式（以及許多其他應用程式）的絕佳選擇。

想了解如何在生產環境中實現這種平行性，請參見 [部屬](deployment/index.md){.internal-link target=_blank}。

## `async` 和 `await`

現代 Python 版本提供一種非常直觀的方式定義非同步程式碼。這使得它看起來就像正常的「順序」程式碼，並在適當的時機「等待」。

當某個操作需要等待才能回傳結果，並且支援這些新的 Python 特性時，你可以像這樣編寫程式碼：

```Python

		if err != nil {
			return nil, err
		}

		if deploymentID == "" {
			// all pools should have same deployment ID
			deploymentID = formats[i].ID
		}

		// Validate if users brought different DeploymentID pools.
		if deploymentID != formats[i].ID {
			return nil, fmt.Errorf("all pools must have same deployment ID - expected %s, got %s for pool(%s)", deploymentID, formats[i].ID, humanize.Ordinal(i+1))
		}

- Promoted ReplicaSet and Deployment `.status.terminatingReplicas` tracking to beta. The `DeploymentReplicaSetTerminatingReplicas` feature gate is now enabled by default. ([#133087](https://github.com/kubernetes/kubernetes/pull/133087), [@atiratree](https://github.com/atiratree)) [SIG...

Le serveur web principal pour exécuter les applications **FastAPI**.

Vous pouvez le combiner avec Gunicorn, pour avoir un serveur multi-processus asynchrone.

Pour plus de détails, consultez la section [Déploiement](deployment/index.md){.internal-link target=_blank}.

///

## Benchmarks et vitesse

		dstRecords, err = globalDNSConfig.Get(dstBucket)
		if err != nil {
			writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
			return
		}

		// Send PutObject request to appropriate instance (in federated deployment)
		core, rerr := getRemoteInstanceClient(r, getHostFromSrv(dstRecords))
		if rerr != nil {
			writeErrorResponse(ctx, w, toAPIError(ctx, rerr), r.URL)
			return
		}

		popts := minio.PutObjectPartOptions{

}

// Returns a minio-go Client configured to access remote host described by destDNSRecord
// Applicable only in a federated deployment
var getRemoteInstanceClient = func(r *http.Request, host string) (*miniogo.Core, error) {
	cred := getReqAccessCred(r, globalSite.Region())
	// In a federated deployment, all the instances share config files
	// and hence expected to have same credentials.

1BC00..1BC6A  ; valid                                  # 7.0  DUPLOYAN LETTER H..DUPLOYAN LETTER VOCALIC M
1BC6B..1BC6F  ; disallowed                             # NA   <reserved-1BC6B>..<reserved-1BC6F>
1BC70..1BC7C  ; valid                                  # 7.0  DUPLOYAN AFFIX LEFT HORIZONTAL SECANT..DUPLOYAN AFFIX ATTACHED TANGENT HOOK

파이썬이 **데이터 사이언스**, 머신러닝과 특히 딥러닝에 의 주된 언어라는 간단한 사실에 더해서, 이것은 FastAPI를 데이터 사이언스 / 머신러닝 웹 API와 응용프로그램에 (다른 것들보다) 좋은 선택지가 되게 합니다.

배포시 병렬을 어떻게 가능하게 하는지 알고싶다면, [배포](deployment/index.md){.internal-link target=_blank}문서를 참고하십시오.

## `async`와  `await`

최신 파이썬 버전에는 비동기 코드를 정의하는 매우 직관적인 방법이 있습니다. 이는 이것을 평범한 "순차적" 코드로 보이게 하고, 적절한 순간에 당신을 위해 "대기"합니다.

- Seccomp support has graduated to GA. A new `seccompProfile` field is added...

- The deprecated kube-controller-manager flag '--deployment-controller-sync-period' has been removed, it is not used by the deployment controller. ([#107178](https://github.com/kubernetes/kubernetes/pull/107178), [@SataQiu](https://github.com/SataQiu))