Skip to content

Commit 0556bd8

Browse files
k8s-ci-robotk8s-ci-robot
authored
Merge pull request #40889 from windsonsea/swiple
[zh] sync switch-to-evented-pleg.md
2 parents 528e0f6 + 8588ced commit 0556bd8

File tree

1 file changed

+187
-0
lines changed

1 file changed

+187
-0
lines changed

‎content/zh-cn/docs/tasks/administer-cluster/switch-to-evented-pleg.md

Lines changed: 187 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,187 @@
1+
---
2+
title: 从轮询切换为基于 CRI 事件的更新来获取容器状态
3+
min-kubernetes-server-version: 1.26
4+
content_type: task
5+
weight: 90
6+
---
7+
<!--
8+
title: Switching from Polling to CRI Event-based Updates to Container Status
9+
min-kubernetes-server-version: 1.26
10+
content_type: task
11+
weight: 90
12+
-->
13+
14+
{{< feature-state for_k8s_version="v1.27" state="beta" >}}
15+
16+
<!-- overview -->
17+
18+
<!--
19+
This page shows how to migrate nodes to use event based updates for container status. The event-based
20+
implementation reduces node resource consumption by the kubelet, compared to the legacy approach
21+
that relies on polling.
22+
You may know this feature as _evented Pod lifecycle event generator (PLEG)_. That's the name used
23+
internally within the Kubernetes project for a key implementation detail.
24+
25+
The polling based approach is referred to as _generic PLEG_.
26+
-->
27+
本页展示了如何迁移节点以使用基于事件的更新来获取容器状态。
28+
与依赖轮询的传统方法相比，基于事件的实现可以减少 kubelet 对节点资源的消耗。
29+
你可以将这个特性称为**事件驱动的 Pod 生命周期事件生成器 (PLEG)**。
30+
这是在 Kubernetes 项目内部针对关键实现细节所用的名称。
31+
32+
基于轮询的方法称为**通用 PLEG**。
33+
34+
## {{% heading "prerequisites" %}}
35+
36+
<!--
37+
* You need to run a version of Kubernetes that provides this feature.
38+
Kubernetes v1.27 includes beta support for event-based container
39+
status updates. The feature is beta but is _disabled_ by default
40+
because it requires support from the container runtime.
41+
* {{< version-check >}}
42+
If you are running a different version of Kubernetes, check the documentation for that release.
43+
-->
44+
* 你需要运行提供此特性的 Kubernetes 版本。
45+
Kubernetes 1.27 提供了对基于事件更新容器状态的 Beta 支持。
46+
此特性处于 Beta 阶段，默认被**禁用**。
47+
* {{< version-check >}}
48+
如果你正在运行不同版本的 Kubernetes，请查阅对应版本的文档。
49+
<!--
50+
* The container runtime in use must support container lifecycle events.
51+
The kubelet automatically switches back to the legacy generic PLEG
52+
mechanism if the container runtime does not announce support for
53+
container lifecycle events, even if you have this feature gate enabled.
54+
-->
55+
* 所使用的容器运行时必须支持容器生命周期事件。
56+
如果容器运行时未声明对容器生命周期事件的支持，即使你已启用了此特性门控，
57+
kubelet 也会自动切换回传统的通用 PLEG。
58+
59+
<!-- steps -->
60+
61+
<!--
62+
## Why switch to Evented PLEG?
63+
64+
* The _Generic PLEG_ incurs non-negligible overhead due to frequent polling of container statuses.
65+
* This overhead is exacerbated by kubelet's parallelized polling of container states, thus limiting
66+
its scalability and causing poor performance and reliability problems.
67+
* The goal of _Evented PLEG_ is to reduce unnecessary work during inactivity
68+
by replacing periodic polling.
69+
-->
70+
## 为什么要切换到事件驱动的 PLEG？ {#why-switch-to-evented-pleg}
71+
72+
* **通用 PLEG** 由于频繁轮询容器状态而产生了不可忽略的开销。
73+
* 这种开销会被 kubelet 的并行轮询容器状态的机制加剧，
74+
限制了可扩缩性，还会导致性能和可靠性问题。
75+
* **事件驱动的 PLEG** 的目标是通过替换定期轮询来减少闲置时的非必要任务。
76+
77+
<!--
78+
## Switching to Evented PLEG
79+
80+
1. Start the kubelet with the [feature gate](/docs/reference/command-line-tools-reference/feature-gates/)
81+
`EventedPLEG` enabled. You can manage the kubelet feature gates editing the kubelet
82+
[config file](/docs/tasks/administer-cluster/kubelet-config-file/) and restarting the kubelet service.
83+
You need to do this on each node where you are using this feature.
84+
-->
85+
## 切换为事件驱动的 PLEG {#switching-to-evented-pleg}
86+
87+
1. 启用[特性门控](/zh-cn/docs/reference/command-line-tools-reference/feature-gates/)
88+
`EventedPLEG` 后启动 kubelet。
89+
你可以通过编辑 kubelet [配置文件](/zh-cn/docs/tasks/administer-cluster/kubelet-config-file/)并重启
90+
kubelet 服务来管理 kubelet 特性门控。
91+
你需要在使用此特性的所有节点上执行此操作。
92+
93+
<!--
94+
2. Make sure the node is [drained](/docs/tasks/administer-cluster/safely-drain-node/) before proceeding.
95+
96+
3. Start the container runtime with the container event generation enabled.
97+
-->
98+
2. 确保节点被[腾空](/zh-cn/docs/tasks/administer-cluster/safely-drain-node/)后再继续。
99+
100+
3. 启用容器事件生成后启动容器运行时。
101+
102+
{{< tabs name="tab_with_code" >}}
103+
104+
{{% tab name="containerd" %}}
105+
<!--
106+
Version 1.7+
107+
-->
108+
版本 1.7+
109+
110+
{{% /tab %}}
111+
112+
{{% tab name="CRI-O" %}}
113+
<!--
114+
Version 1.26+
115+
116+
Check if the CRI-O is already configured to emit CRI events by verifying the configuration,
117+
-->
118+
版本 1.26+
119+
120+
通过验证配置，检查 CRI-O 是否已配置为发送 CRI 事件：
121+
122+
```shell
123+
crio config | grep enable_pod_events
124+
```
125+
126+
<!--
127+
If it is enabled, the output should be similar to the following:
128+
-->
129+
如果已启用，输出应类似于：
130+
131+
```none
132+
enable_pod_events = true
133+
```
134+
135+
<!--
136+
To enable it, start the CRI-O daemon with the flag `--enable-pod-events=true` or
137+
use a dropin config with the following lines:
138+
-->
139+
要启用它，可使用 `--enable-pod-events=true` 标志或添加以下配置来启动 CRI-O 守护进程：
140+
141+
```toml
142+
[crio.runtime]
143+
enable_pod_events: true
144+
```
145+
146+
{{% /tab %}}
147+
{{< /tabs >}}
148+
149+
{{< version-check >}}
150+
151+
<!--
152+
4. Verify that the kubelet is using event-based container stage change monitoring.
153+
To check, look for the term `EventedPLEG` in the kubelet logs.
154+
155+
The output should be similar to this:
156+
-->
157+
4. 确认 kubelet 正使用基于事件的容器阶段变更监控。
158+
要检查这一点，可在 kubelet 日志中查找 `EventedPLEG` 词条。
159+
160+
输出类似于：
161+
162+
```console
163+
I0314 11:10:13.909915 1105457 feature_gate.go:249] feature gates: &{map[EventedPLEG:true]}
164+
```
165+
166+
<!--
167+
If you have set `--v` to 4 and above, you might see more entries that indicate
168+
that the kubelet is using event-based container state monitoring.
169+
-->
170+
如果你将 `--v` 设置为 4 及更高值，你可能会看到更多条目表明
171+
kubelet 正在使用基于事件的容器状态监控。
172+
173+
```console
174+
I0314 11:12:42.009542 1110177 evented.go:238] "Evented PLEG: Generated pod status from the received event" podUID=3b2c6172-b112-447a-ba96-94e7022912dc
175+
I0314 11:12:44.623326 1110177 evented.go:238] "Evented PLEG: Generated pod status from the received event" podUID=b3fba5ea-a8c5-4b76-8f43-481e17e8ec40
176+
I0314 11:12:44.714564 1110177 evented.go:238] "Evented PLEG: Generated pod status from the received event" podUID=b3fba5ea-a8c5-4b76-8f43-481e17e8ec40
177+
```
178+
179+
## {{% heading "whatsnext" %}}
180+
181+
<!--
182+
* Learn more about the design in the Kubernetes Enhancement Proposal (KEP):
183+
[kubelet Evented PLEG for Better Performance](https://github.com/kubernetes/enhancements/blob/5b258a990adabc2ffdc9d84581ea6ed696f7ce6c/keps/sig-node/3386-kubelet-evented-pleg/README.md).
184+
-->
185+
* 进一步了解 Kubernetes 增强提案 (KEP)：
186+
[kubelet Evented PLEG for Better Performance](https://github.com/kubernetes/enhancements/blob/5b258a990adabc2ffdc9d84581ea6ed696f7ce6c/keps/sig-node/3386-kubelet-evented-pleg/README.md)
187+
中的设计理念。

0 commit comments

Comments
 (0)