Mit **FastAPI** können Sie die Vorteile der Nebenläufigkeit nutzen, die in der Webentwicklung weit verbreitet ist (derselbe Hauptvorteil von NodeJS).

Sie können aber auch die Vorteile von Parallelität und Multiprocessing (mehrere Prozesse werden parallel ausgeführt) für **CPU-lastige** Workloads wie in Systemen für maschinelles Lernen nutzen.

* Return apiserver panics as 500 errors instead terminating the apiserver process. ([#68001](https://github.com/kubernetes/kubernetes/pull/68001), [@sttts](https://github.com/sttts))

* Local volume source paths that are directories and not mount points fail to unmount.  A fix is in process ([#48331](https://github.com/kubernetes/kubernetes/issues/48331)).

Starlette 是一個輕量的 <dfn title="用於構建非同步 Python 網頁應用的新標準">ASGI</dfn> 框架／工具集，非常適合用來建構高效能的 asyncio 服務。

它非常簡單直觀。設計上易於擴充，且元件化。

它具備：

* 令人印象深刻的效能。
* WebSocket 支援。
* 行程內（in-process）背景任務。
* 啟動與關閉事件。
* 建立在 HTTPX 上的測試用戶端。
* CORS、GZip、靜態檔案、串流回應。
* Session 與 Cookie 支援。
* 100% 測試涵蓋率。
* 100% 型別註解的程式碼庫。
* 幾乎沒有硬性相依。

Starlette 目前是測試中最快的 Python 框架。僅次於 Uvicorn（它不是框架，而是伺服器）。

          <name>profiles</name>
          <version>4.0.0+</version>
          <description>A listing of project-local build profiles which will modify the build process
            when activated.</description>
          <association>
            <type>Profile</type>
            <multiplicity>*</multiplicity>
          </association>
        </field>
      </fields>

for purposes such as debugging, similar to how `kubectl exec` runs a process in an existing container. Also like `kubectl exec`, no resources are reserved for ephemeral containers and they are not restarted when they exit. Note that container namespace targeting is not yet implemented, so [process namespace sharing](https://kubernetes.io/docs/tasks/configure-pod-container/share-process-namespace/) must be enabled to view process from other containers in the pod. ([#59484](https://github.com/kuber...

	}
	word, cond = p.lex.Text(), ""
	operands = scratch[:0]
	// Zero or more comma-separated operands, one per loop.
	nesting := 0
	colon := -1
	for tok != '\n' && tok != ';' {
		// Process one operand.
		var items []lex.Token
		if cap(operands) > len(operands) {
			// Reuse scratch items slice.
			items = operands[:cap(operands)][len(operands)][:0]
		} else {
			items = make([]lex.Token, 0, 3)

- Added a redundant process to remove tracking finalizers from Pods that belong to Jobs. The process kicks in after the control plane marks a Job as finished. ([#119944](https://github.com/kubernetes/kubernetes/pull/119944), [@Sharpz7](https://github.com/Sharpz7))

		}, index)
	}

	for index, err := range g.Wait() {
		if err != nil {
			sys.addNodeErr(&reply[index], sys.peerClients[index], err)
		}
	}
	return reply
}

// GetProcInfo - Process information
func (sys *NotificationSys) GetProcInfo(ctx context.Context) []madmin.ProcInfo {
	reply := make([]madmin.ProcInfo, len(sys.peerClients))

	g := errgroup.WithNErrs(len(sys.peerClients))

Multi-arch support (EXPERIMENTAL):
  - kubeadm now adds support for docker “schema 2” manifest lists. This is preliminary part of the process of making kubeadm based k8s deployments to support multiple architectures.
Deprecating features:
  - The Alpha feature-gates HighAvailability, SelfHosting, CertsInSecrets are now deprecated, and will be removed in k8s v1.13.0.