* Fixed an issue reporting lack of progress for a deployment prematurely ([#52178](https://github.com/kubernetes/kubernetes/pull/52178), [@kargakis](https://github.com/kargakis))
* [fluentd-gcp addon] Bug with event-exporter leaking memory on metrics in clusters with CA is fixed. ([#52263](https://github.com/kubernetes/kubernetes/pull/52263), [@crassirostris](https://github.com/crassirostris))



# v1.7.6

 *
 * @author Kevin Bourrillion
 * @author Nishant Thakkar
 * @author Sven Mawson
 * @since 1.0
 */
@GwtCompatible
public final class Futures extends GwtFuturesCatchingSpecialization {

  // A note on memory visibility.
  // Many of the utilities in this class (transform, withFallback, withTimeout, asList, combine)
  // have two requirements that significantly complicate their design.

 * your HTTP calls. This is because each client holds its own connection pool and thread pools.
 * Reusing connections and threads reduces latency and saves memory. Conversely, creating a client
 * for each request wastes resources on idle pools.
 *
 * Use `new OkHttpClient()` to create a shared instance with the default settings:
 *
 * ```java
 * // The singleton HTTP client.

* Add `metrics-port` to kube-proxy cmd flags. ([#72682](https://github.com/kubernetes/kubernetes/pull/72682), [@whypro](https://github.com/whypro))
* Reduce memory utilization of admission webhook metrics by removing resource related labels. ([#69895](https://github.com/kubernetes/kubernetes/pull/69895), [@jpbetz](https://github.com/jpbetz))



# v1.11.8

Ardından `yield`’den hemen sonra modeli bellekten kaldırıyoruz (unload). Bu kod, uygulama **request’leri işlemeyi bitirdikten sonra**, *shutdown*’dan hemen önce çalıştırılır. Örneğin memory veya GPU gibi resource’ları serbest bırakabilir.

/// tip | İpucu

`shutdown`, uygulamayı **durdurduğunuzda** gerçekleşir.

Belki yeni bir sürüm başlatmanız gerekiyordur, ya da çalıştırmaktan sıkılmışsınızdır. 🤷

- Fixes a kubelet and kube-apiserver memory leak in default 1.29 configurations related to tracing. ([#126983](https://github.com/kubernetes/kubernetes/pull/126983), [@dashpole](https://github.com/dashpole)) [SIG API Machinery and Node]

- Fixes a kubelet and kube-apiserver memory leak in default 1.29 configurations related to tracing. ([#126984](https://github.com/kubernetes/kubernetes/pull/126984), [@dashpole](https://github.com/dashpole)) [SIG API Machinery and Node]

- Fixed a bug where startupProbe stopped working after a container's first restart ([#101093](https://github.com/kubernetes/kubernetes/pull/101093), [@wzshiming](https://github.com/wzshiming)) [SIG Node]
- Fixed port-forward memory leak for long-running and heavily used connections. ([#99839](https://github.com/kubernetes/kubernetes/pull/99839), [@saschagrunert](https://github.com/saschagrunert)) [SIG API Machinery and Node]

那你可能會想要「單一容器」搭配「行程管理器」，在其中啟動「多個 worker 行程」。

---

重點是，這些「都不是」必須盲目遵守的「鐵律」。你可以用這些想法來「評估你的使用情境」，並決定對你的系統最好的做法，看看如何管理以下概念：

* 安全性 - HTTPS
* 開機自動執行
* 失敗重啟
* 複本（執行的行程數量）
* 記憶體
* 啟動前的前置步驟

## 記憶體 { #memory }

如果你採用「每個容器一個行程」，那每個容器（若有複本則多個容器）所消耗的記憶體會是相對明確、穩定且有限的。

接著你可以在容器管理系統（例如 Kubernetes）的設定中為容器設定相同的記憶體限制與需求。如此，它就能在「可用的機器」上「複製容器」，並考量容器所需的記憶體量與叢集中機器的可用記憶體。

    Node Allocatable. This will result in a reduction of schedulable capacity in
    clusters post upgrade where kubelet hard eviction has been turned on for
    memory. To opt-out set `--experimental-allocatable-ignore-eviction=true`.
  * More details on these feature here:
    https://kubernetes.io/docs/tasks/administer-cluster/reserve-compute-resources/