@@ -177,7 +177,7 @@ description: "此优先级类应仅用于 XYZ 服务 Pod。"
177
177
178
178
{{< feature-state for_k8s_version="v1.19" state="beta" >}}
179
179
180
- Pods with ` PreemptionPolicy : Never` will be placed in the scheduling queue
180
+ Pods with ` preemptionPolicy : Never` will be placed in the scheduling queue
181
181
ahead of lower-priority pods,
182
182
but they cannot preempt other pods.
183
183
A non-preempting pod waiting to be scheduled will stay in the scheduling queue,
@@ -197,7 +197,7 @@ high-priority pods.
197
197
198
198
{{< feature-state for_k8s_version="v1.19" state="beta" >}}
199
199
200
- 配置了 `PreemptionPolicy : Never` 的 Pod 将被放置在调度队列中较低优先级 Pod 之前,
200
+ 配置了 `preemptionPolicy : Never` 的 Pod 将被放置在调度队列中较低优先级 Pod 之前,
201
201
但它们不能抢占其他 Pod。等待调度的非抢占式 Pod 将留在调度队列中,直到有足够的可用资源,
202
202
它才可以被调度。非抢占式 Pod,像其他 Pod 一样,受调度程序回退的影响。
203
203
这意味着如果调度程序尝试这些 Pod 并且无法调度它们,它们将以更低的频率被重试,
@@ -206,26 +206,26 @@ high-priority pods.
206
206
非抢占式 Pod 仍可能被其他高优先级 Pod 抢占。
207
207
208
208
<!--
209
- ` PreemptionPolicy ` defaults to `PreemptLowerPriority`,
209
+ ` preemptionPolicy ` defaults to `PreemptLowerPriority`,
210
210
which will allow pods of that PriorityClass to preempt lower-priority pods
211
211
(as is existing default behavior).
212
- If `PreemptionPolicy ` is set to `Never`,
212
+ If `preemptionPolicy ` is set to `Never`,
213
213
pods in that PriorityClass will be non-preempting.
214
214
215
215
An example use case is for data science workloads.
216
216
A user may submit a job that they want to be prioritized above other workloads,
217
217
but do not wish to discard existing work by preempting running pods.
218
- The high priority job with `PreemptionPolicy : Never` will be scheduled
218
+ The high priority job with `preemptionPolicy : Never` will be scheduled
219
219
ahead of other queued pods,
220
220
as soon as sufficient cluster resources "naturally" become free.
221
221
-->
222
- ` PreemptionPolicy ` 默认为 `PreemptLowerPriority`,
222
+ ` preemptionPolicy ` 默认为 `PreemptLowerPriority`,
223
223
这将允许该 PriorityClass 的 Pod 抢占较低优先级的 Pod(现有默认行为也是如此)。
224
- 如果 `PreemptionPolicy ` 设置为 `Never`,则该 PriorityClass 中的 Pod 将是非抢占式的。
224
+ 如果 `preemptionPolicy ` 设置为 `Never`,则该 PriorityClass 中的 Pod 将是非抢占式的。
225
225
226
226
数据科学工作负载是一个示例用例。用户可以提交他们希望优先于其他工作负载的作业,
227
227
但不希望因为抢占运行中的 Pod 而导致现有工作被丢弃。
228
- 设置为 `PreemptionPolicy : Never` 的高优先级作业将在其他排队的 Pod 之前被调度,
228
+ 设置为 `preemptionPolicy : Never` 的高优先级作业将在其他排队的 Pod 之前被调度,
229
229
只要足够的集群资源“自然地”变得可用。
230
230
231
231
<!-- # ## Example Non-preempting PriorityClass -->
@@ -664,4 +664,4 @@ kubelet 使用优先级来确定
664
664
[默认限制优先级消费](/zh/docs/concepts/policy/resource-quotas/#limit-priority-class-consumption-by-default)
665
665
* 了解 [Pod 干扰](/zh/docs/concepts/workloads/pods/disruptions/)
666
666
* 了解 [API 发起的驱逐](/zh/docs/concepts/scheduling-eviction/api-eviction/)
667
- * 了解[节点压力驱逐](/zh/docs/concepts/scheduling-eviction/node-pressure-eviction/)
667
+ * 了解[节点压力驱逐](/zh/docs/concepts/scheduling-eviction/node-pressure-eviction/)
0 commit comments