Skip to content

Commit 62e54e5

Browse files
k8s-ci-robotk8s-ci-robot
authored
Merge pull request #45206 from lbzss/autoscaling.md
[zh-cn] Localize autoscaling.md
2 parents 4339ba7 + af0fcc8 commit 62e54e5

File tree

2 files changed

+350
-0
lines changed

2 files changed

+350
-0
lines changed

‎content/zh-cn/docs/concepts/workloads/autoscaling.md

Lines changed: 297 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,297 @@
1+
---
2+
title: 自动扩缩工作负载
3+
description: >-
4+
通过自动扩缩，你可以用某种方式自动更新你的工作负载。在面对资源需求变化的时候可以使你的集群更灵活、更高效。
5+
content_type: concept
6+
weight: 40
7+
---
8+
<!--
9+
title: Autoscaling Workloads
10+
description: >-
11+
With autoscaling, you can automatically update your workloads in one way or another. This allows your cluster to react to changes in resource demand more elastically and efficiently.
12+
content_type: concept
13+
weight: 40
14+
-->
15+
16+
<!-- overview -->
17+
18+
<!--
19+
In Kubernetes, you can _scale_ a workload depending on the current demand of resources.
20+
This allows your cluster to react to changes in resource demand more elastically and efficiently.
21+
22+
When you scale a workload, you can either increase or decrease the number of replicas managed by
23+
the workload, or adjust the resources available to the replicas in-place.
24+
25+
The first approach is referred to as _horizontal scaling_, while the second is referred to as
26+
_vertical scaling_.
27+
28+
There are manual and automatic ways to scale your workloads, depending on your use case.
29+
-->
30+
在 Kubernetes 中，你可以根据当前的资源需求**扩缩**工作负载。
31+
这让你的集群可以更灵活、更高效地面对资源需求的变化。
32+
33+
当你扩缩工作负载时，你可以增加或减少工作负载所管理的副本数量，或者就地调整副本的可用资源。
34+
35+
第一种手段称为**水平扩缩**，第二种称为**垂直扩缩**。
36+
37+
扩缩工作负载有手动和自动两种方式，这取决于你的使用情况。
38+
39+
<!-- body -->
40+
41+
<!--
42+
## Scaling workloads manually
43+
-->
44+
## 手动扩缩工作负载 {#scaling-workloads-manually}
45+
46+
<!--
47+
Kubernetes supports _manual scaling_ of workloads. Horizontal scaling can be done
48+
using the `kubectl` CLI.
49+
For vertical scaling, you need to _patch_ the resource definition of your workload.
50+
51+
See below for examples of both strategies.
52+
-->
53+
Kubernetes 支持工作负载的手动扩缩。水平扩缩可以使用 `kubectl` 命令行工具完成。
54+
对于垂直扩缩，你需要**更新**工作负载的资源定义。
55+
56+
这两种策略的示例见下文。
57+
58+
<!--
59+
- **Horizontal scaling**: [Running multiple instances of your app](/docs/tutorials/kubernetes-basics/scale/scale-intro/)
60+
- **Vertical scaling**: [Resizing CPU and memory resources assigned to containers](/docs/tasks/configure-pod-container/resize-container-resources)
61+
-->
62+
- **水平扩缩**：[运行应用程序的多个实例](/docs/tutorials/kubernetes-basics/scale/scale-intro/)
63+
- **垂直扩缩**：[调整分配给容器的 CPU 和内存资源](/docs/tasks/configure-pod-container/resize-container-resources)
64+
65+
<!--
66+
## Scaling workloads automatically
67+
-->
68+
## 自动扩缩工作负载 {#scaling-workloads-automatically}
69+
70+
<!--
71+
Kubernetes also supports _automatic scaling_ of workloads, which is the focus of this page.
72+
-->
73+
Kubernetes 也支持工作负载的**自动扩缩**，这也是本页的重点。
74+
75+
<!--
76+
The concept of _Autoscaling_ in Kubernetes refers to the ability to automatically update an
77+
object that manages a set of Pods (for example a
78+
{{< glossary_tooltip text="Deployment" term_id="deployment" >}}.
79+
-->
80+
在 Kubernetes 中**自动扩缩**的概念是指自动更新管理一组 Pod 的能力（例如
81+
{{< glossary_tooltip text="Deployment" term_id="deployment" >}}）。
82+
83+
<!--
84+
### Scaling workloads horizontally
85+
-->
86+
### 水平扩缩工作负载 {#scaling-workloads-horizontally}
87+
88+
<!--
89+
In Kubernetes, you can automatically scale a workload horizontally using a _HorizontalPodAutoscaler_ (HPA).
90+
-->
91+
在 Kubernetes 中，你可以使用 HorizontalPodAutoscaler (HPA) 实现工作负载的自动水平扩缩。
92+
93+
<!--
94+
It is implemented as a Kubernetes API resource and a {{< glossary_tooltip text="controller" term_id="controller" >}}
95+
and periodically adjusts the number of {{< glossary_tooltip text="replicas" term_id="replica" >}}
96+
in a workload to match observed resource utilization such as CPU or memory usage.
97+
-->
98+
它以 Kubernetes API 资源和{{< glossary_tooltip text="控制器" term_id="controller" >}}的方式实现，
99+
并定期调整工作负载中{{< glossary_tooltip text="副本" term_id="replica" >}}的数量
100+
以满足设置的资源利用率，如 CPU 或内存利用率。
101+
102+
<!--
103+
There is a [walkthrough tutorial](/docs/tasks/run-application/horizontal-pod-autoscale-walkthrough) of configuring a HorizontalPodAutoscaler for a Deployment.
104+
-->
105+
这是一个为 Deployment 部署配置 HorizontalPodAutoscaler 的[示例教程](/docs/tasks/run-application/horizontal-pod-autoscale-walkthrough)。
106+
107+
<!--
108+
### Scaling workloads vertically
109+
-->
110+
### 垂直扩缩工作负载 {#scaling-workloads-vertically}
111+
112+
{{< feature-state for_k8s_version="v1.25" state="stable" >}}
113+
114+
<!--
115+
You can automatically scale a workload vertically using a _VerticalPodAutoscaler_ (VPA).
116+
Different to the HPA, the VPA doesn't come with Kubernetes by default, but is a separate project
117+
that can be found [on GitHub](https://github.com/kubernetes/autoscaler/tree/9f87b78df0f1d6e142234bb32e8acbd71295585a/vertical-pod-autoscaler).
118+
-->
119+
你可以使用 VerticalPodAutoscaler (VPA) 实现工作负载的垂直扩缩。
120+
不同于 HPA，VPA 并非默认来源于 Kubernetes，而是一个独立的项目，
121+
参见 [on GitHub](https://github.com/kubernetes/autoscaler/tree/9f87b78df0f1d6e142234bb32e8acbd71295585a/vertical-pod-autoscaler)。
122+
123+
<!--
124+
Once installed, it allows you to create {{< glossary_tooltip text="CustomResourceDefinitions" term_id="customresourcedefinition" >}}
125+
(CRDs) for your workloads which define _how_ and _when_ to scale the resources of the managed replicas.
126+
-->
127+
安装后，你可以为工作负载创建 {{< glossary_tooltip text="CustomResourceDefinitions" term_id="customresourcedefinition" >}}(CRDs)，
128+
定义**如何**以及**何时**扩缩被管理副本的资源。
129+
130+
{{< note >}}
131+
<!--
132+
You will need to have the [Metrics Server](https://github.com/kubernetes-sigs/metrics-server)
133+
installed to your cluster for the HPA to work.
134+
-->
135+
你需要在集群中安装 [Metrics Server](https://github.com/kubernetes-sigs/metrics-server)，这样，你的 HPA 才能正常工作。
136+
{{< /note >}}
137+
138+
<!--
139+
At the moment, the VPA can operate in four different modes:
140+
-->
141+
目前，VPA 可以有四种不同的运行模式：
142+
143+
<!--
144+
{{< table caption="Different modes of the VPA" >}}
145+
Mode | Description
146+
:----|:-----------
147+
`Auto` | Currently `Recreate`, might change to in-place updates in the future
148+
`Recreate` | The VPA assigns resource requests on pod creation as well as updates them on existing pods by evicting them when the requested resources differ significantly from the new recommendation
149+
`Initial` | The VPA only assigns resource requests on pod creation and never changes them later.
150+
`Off` | The VPA does not automatically change the resource requirements of the pods. The recommendations are calculated and can be inspected in the VPA object.
151+
{{< /table >}}
152+
-->
153+
{{< table caption="VPA 的不同模式" >}}
154+
模式 | 描述
155+
:----|:-----------
156+
`Auto` | 目前是 `Recreate`，将来可能改为就地更新
157+
`Recreate` | VPA 会在创建 Pod 时分配资源请求，并且当请求的资源与新的建议值区别很大时通过驱逐 Pod 的方式来更新现存的 Pod
158+
`Initial` | VPA 只有在创建时分配资源请求，之后不做更改
159+
`Off` | VPA 不会自动更改 Pod 的资源需求，建议值仍会计算并可在 VPA 对象中查看
160+
{{< /table >}}
161+
162+
<!--
163+
#### Requirements for in-place resizing
164+
-->
165+
#### 就地调整的要求
166+
167+
{{< feature-state for_k8s_version="v1.27" state="alpha" >}}
168+
169+
<!--
170+
Resizing a workload in-place **without** restarting the {{< glossary_tooltip text="Pods" term_id="pod" >}}
171+
or its {{< glossary_tooltip text="Containers" term_id="container" >}} requires Kubernetes version 1.27 or later.<br />
172+
Additionally, the `InPlaceVerticalScaling` feature gate needs to be enabled.
173+
-->
174+
在**不**重启 {{< glossary_tooltip text="Pod" term_id="pod" >}} 或其中{{< glossary_tooltip text="容器" term_id="container" >}}就地调整工作负载的情况下要求 Kubernetes 版本大于 1.27。
175+
此外，特性门控 `InPlaceVerticalScaling` 需要开启。
176+
177+
{{< feature-gate-description name="InPlacePodVerticalScaling" >}}
178+
179+
<!--
180+
### Autoscaling based on cluster size
181+
-->
182+
### 根据集群规模自动扩缩 {#autoscaling-based-on-cluster-size}
183+
184+
<!--
185+
For workloads that need to be scaled based on the size of the cluster (for example
186+
`cluster-dns` or other system components), you can use the
187+
[_Cluster Proportional Autoscaler_](https://github.com/kubernetes-sigs/cluster-proportional-autoscaler).<br />
188+
Just like the VPA, it is not part of the Kubernetes core, but hosted as its
189+
own project on GitHub.
190+
-->
191+
对于需要根据集群规模实现扩缩的工作负载（例如：`cluster-dns` 或者其他系统组件），
192+
你可以使用 [Cluster Proportional Autoscaler](https://github.com/kubernetes-sigs/cluster-proportional-autoscaler)。
193+
与 VPA 一样，这个项目不是 Kubernetes 核心项目的一部分，它在 GitHub 上有自己的项目。
194+
195+
<!--
196+
The Cluster Proportional Autoscaler watches the number of schedulable {{< glossary_tooltip text="nodes" term_id="node" >}}
197+
and cores and scales the number of replicas of the target workload accordingly.
198+
-->
199+
集群弹性伸缩器 (Cluster Proportional Autoscaler) 会观测可调度 {{< glossary_tooltip text="节点" term_id="node" >}} 和 内核数量，
200+
并调整目标工作负载的副本数量。
201+
202+
<!--
203+
If the number of replicas should stay the same, you can scale your workloads vertically according to the cluster size using
204+
the [_Cluster Proportional Vertical Autoscaler_](https://github.com/kubernetes-sigs/cluster-proportional-vertical-autoscaler).
205+
The project is **currently in beta** and can be found on GitHub.
206+
-->
207+
如果副本的数量需要保持一致，你可以使用 [Cluster Proportional Vertical Autoscaler](https://github.com/kubernetes-sigs/cluster-proportional-vertical-autoscaler) 来根据集群规模进行垂直扩缩。
208+
这个项目目前处于 **beta** 阶段，你可以在 GitHub 上找到它。
209+
210+
<!--
211+
While the Cluster Proportional Autoscaler scales the number of replicas of a workload, the Cluster Proportional Vertical Autoscaler
212+
adjusts the resource requests for a workload (for example a Deployment or DaemonSet) based on the number of nodes and/or cores
213+
in the cluster.
214+
-->
215+
集群弹性伸缩器会扩缩工作负载的副本数量，垂直集群弹性伸缩器 (Cluster Proportional Vertical Autoscaler) 会根据节点和/或核心的数量
216+
调整工作负载的资源请求（例如 Deployment 和 DaemonSet）。
217+
218+
<!--
219+
### Event driven Autoscaling
220+
-->
221+
### 事件驱动型自动扩缩 {#event-driven-autoscaling}
222+
223+
<!--
224+
It is also possible to scale workloads based on events, for example using the
225+
[_Kubernetes Event Driven Autoscaler_ (**KEDA**)](https://keda.sh/).
226+
-->
227+
通过事件驱动实现工作负载的扩缩也是可行的，
228+
例如使用 [Kubernetes Event Driven Autoscaler (**KEDA**)](https://keda.sh/)。
229+
230+
<!--
231+
KEDA is a CNCF graduated enabling you to scale your workloads based on the number
232+
of events to be processed, for example the amount of messages in a queue. There exists
233+
a wide range of adapters for different event sources to choose from.
234+
-->
235+
KEDA 是 CNCF 的毕业项目，能让你根据要处理事件的数量对工作负载进行扩缩，例如队列中消息的数量。
236+
有多种针对不同事件源的适配可供选择。
237+
238+
<!--
239+
### Autoscaling based on schedules
240+
-->
241+
### 根据计划自动扩缩 {#autoscaling-based-on-schedules}
242+
243+
<!--
244+
Another strategy for scaling your workloads is to **schedule** the scaling operations, for example in order to
245+
reduce resource consumption during off-peak hours.
246+
-->
247+
扩缩工作负载的另一种策略是**计划**进行扩缩，例如在非高峰时段减少资源消耗。
248+
249+
<!--
250+
Similar to event driven autoscaling, such behavior can be achieved using KEDA in conjunction with
251+
its [`Cron` scaler](https://keda.sh/docs/2.13/scalers/cron/). The `Cron` scaler allows you to define schedules
252+
(and time zones) for scaling your workloads in or out.
253+
-->
254+
与事件驱动型自动扩缩相似，这种行为可以使用 KEDA 和 [`Cron` scaler](https://keda.sh/docs/2.13/scalers/cron/) 实现。
255+
你可以在计划扩缩器 (Cron scaler) 中定义计划来实现工作负载的横向扩缩。
256+
257+
<!--
258+
## Scaling cluster infrastructure
259+
-->
260+
## 扩缩集群基础设施 {#scaling-cluster-infrastructure}
261+
262+
<!--
263+
If scaling workloads isn't enough to meet your needs, you can also scale your cluster infrastructure itself.
264+
-->
265+
如果扩缩工作负载无法满足你的需求，你也可以扩缩集群基础设施本身。
266+
267+
<!--
268+
Scaling the cluster infrastructure normally means adding or removing {{< glossary_tooltip text="nodes" term_id="node" >}}.
269+
This can be done using one of two available autoscalers:
270+
-->
271+
扩缩集群基础设施通常是指增加或移除{{< glossary_tooltip text="节点" term_id="node" >}}。
272+
这可以通过以下两种自动扩缩器中的任意一种实现：
273+
274+
- [**Cluster Autoscaler**](https://github.com/kubernetes/autoscaler/tree/master/cluster-autoscaler)
275+
- [**Karpenter**](https://github.com/kubernetes-sigs/karpenter?tab=readme-ov-file)
276+
277+
<!--
278+
Both scalers work by watching for pods marked as _unschedulable_ or _underutilized_ nodes and then adding or
279+
removing nodes as needed.
280+
-->
281+
这两种扩缩器的工作原理都是通过监测节点上被标记为 **unschedulable** 或 **underutilized** 的 Pod 数量，
282+
然后根据需要增加或移除节点。
283+
284+
## {{% heading "whatsnext" %}}
285+
286+
<!--
287+
- Learn more about scaling horizontally
288+
- [Scale a StatefulSet](/docs/tasks/run-application/scale-stateful-set/)
289+
- [HorizontalPodAutoscaler Walkthrough](/docs/tasks/run-application/horizontal-pod-autoscale-walkthrough/)
290+
- [Resize Container Resources In-Place](/docs/tasks/configure-pod-container/resize-container-resources/)
291+
- [Autoscale the DNS Service in a Cluster](/docs/tasks/administer-cluster/dns-horizontal-autoscaling/)
292+
-->
293+
- 了解有关横向扩缩的更多信息
294+
- [扩缩 StatefulSet](/zh-cn/docs/tasks/run-application/scale-stateful-set/)
295+
- [HorizontalPodAutoscaler 演练](/zh-cn/docs/tasks/run-application/horizontal-pod-autoscale-walkthrough/)
296+
- [调整分配给容器的 CPU 和内存资源](/zh-cn/docs/tasks/configure-pod-container/resize-container-resources/)
297+
- [自动扩缩集群 DNS 服务](/zh-cn/docs/tasks/administer-cluster/dns-horizontal-autoscaling/)

‎content/zh-cn/docs/reference/glossary/replica.md

Lines changed: 53 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,53 @@
1+
---
2+
title: 副本（Replica）
3+
id: replica
4+
date: 2023-06-11
5+
full_link:
6+
short_description: >
7+
Replicas 是 Pod 的副本，通过维护相同的实例确保可用性、可扩缩性和容错性。
8+
aka:
9+
tags:
10+
- fundamental
11+
- workload
12+
---
13+
14+
<!--
15+
title: Replica
16+
id: replica
17+
date: 2023-06-11
18+
full_link:
19+
short_description: >
20+
Replicas are copies of pods, ensuring availability, scalability, and fault tolerance by maintaining identical instances.
21+
aka:
22+
tags:
23+
- fundamental
24+
- workload
25+
-->
26+
27+
<!--
28+
A copy or duplicate of a {{< glossary_tooltip text="Pod" term_id="pod" >}} or
29+
a set of pods. Replicas ensure high availability, scalability, and fault tolerance
30+
by maintaining multiple identical instances of a pod.
31+
-->
32+
单个 {{< glossary_tooltip text="Pod" term_id="pod" >}} 或一组 Pod 的复制拷贝。
33+
Replicas 通过维护多个相同的 Pod 实例保证了高可用性、可扩缩性和容错性。
34+
35+
<!--more-->
36+
<!--
37+
Replicas are commonly used in Kubernetes to achieve the desired application state and reliability.
38+
They enable workload scaling and distribution across multiple nodes in a cluster.
39+
40+
By defining the number of replicas in a Deployment or ReplicaSet, Kubernetes ensures that
41+
the specified number of instances are running, automatically adjusting the count as needed.
42+
43+
Replica management allows for efficient load balancing, rolling updates, and
44+
self-healing capabilities in a Kubernetes cluster.
45+
-->
46+
Kubernetes 中通常使用副本来实现期望的应用状态和可靠性。
47+
它们可以在集群的多个节点上扩缩和分配工作负载。
48+
49+
在 Deployment 或 ReplicaSet 中定义副本数量， Kubernetes 确保了所期望数量的实例正在运行，
50+
并且会根据需要自动调整这个数量。
51+
52+
副本管理可以在 Kubernetes 集群中提供了高效的负载均衡、滚动更新和自愈能力。
53+

0 commit comments

Comments
 (0)