@@ -8,7 +8,6 @@ weight: 20
88reviewers:
99- davidopp
1010- kevin-wangzefeng
11- - bsalamat
1211- alculquicondor
1312title: Assigning Pods to Nodes
1413content_type: concept
@@ -18,13 +17,13 @@ weight: 20
1817<!-- overview -->
1918
2019<!--
21- You can constrain a {{< glossary_tooltip text="Pod" term_id="pod" >}} so that it is
20+ You can constrain a {{< glossary_tooltip text="Pod" term_id="pod" >}} so that it is
2221_restricted_ to run on particular {{< glossary_tooltip text="node(s)" term_id="node" >}},
2322or to _prefer_ to run on particular nodes.
2423There are several ways to do this and the recommended approaches all use
2524[label selectors](/docs/concepts/overview/working-with-objects/labels/) to facilitate the selection.
2625Often, you do not need to set any such constraints; the
27- {{< glossary_tooltip text="scheduler" term_id="kube-scheduler" >}} will automatically do a reasonable placement
26+ {{< glossary_tooltip text="scheduler" term_id="kube-scheduler" >}} will automatically do a reasonable placement
2827(for example, spreading your Pods across nodes so as not place Pods on a node with insufficient free resources).
2928However, there are some circumstances where you may want to control which node
3029the Pod deploys to, for example, to ensure that a Pod ends up on a node with an SSD attached to it,
@@ -46,10 +45,10 @@ Pod 被部署到哪个节点。例如，确保 Pod 最终落在连接了 SSD 的
4645You can use any of the following methods to choose where Kubernetes schedules
4746specific Pods:
4847
49- * [nodeSelector](#nodeselector) field matching against [node labels](#built-in-node-labels)
50- * [Affinity and anti-affinity](#affinity-and-anti-affinity)
51- * [nodeName](#nodename) field
52- * [Pod topology spread constraints](#pod-topology-spread-constraints)
48+ * [nodeSelector](#nodeselector) field matching against [node labels](#built-in-node-labels)
49+ * [Affinity and anti-affinity](#affinity-and-anti-affinity)
50+ * [nodeName](#nodename) field
51+ * [Pod topology spread constraints](#pod-topology-spread-constraints)
5352-->
5453你可以使用下列方法中的任何一种来选择 Kubernetes 对特定 Pod 的调度：
5554
@@ -90,7 +89,7 @@ and a different value in other environments.
9089Adding labels to nodes allows you to target Pods for scheduling on specific
9190nodes or groups of nodes. You can use this functionality to ensure that specific
9291Pods only run on nodes with certain isolation, security, or regulatory
93- properties.
92+ properties.
9493-->
9594## 节点隔离/限制 {#node-isolation-restriction}
9695
@@ -110,7 +109,7 @@ itself so that the scheduler schedules workloads onto the compromised node.
110109<!--
111110The [`NodeRestriction` admission plugin](/docs/reference/access-authn-authz/admission-controllers/#noderestriction)
112111prevents the kubelet from setting or modifying labels with a
113- `node-restriction.kubernetes.io/` prefix.
112+ `node-restriction.kubernetes.io/` prefix.
114113
115114To make use of that label prefix for node isolation:
116115-->
@@ -138,7 +137,7 @@ kubelet 使用 `node-restriction.kubernetes.io/` 前缀设置或修改标签。
138137You can add the `nodeSelector` field to your Pod specification and specify the
139138[node labels](#built-in-node-labels) you want the target node to have.
140139Kubernetes only schedules the Pod onto nodes that have each of the labels you
141- specify.
140+ specify.
142141-->
143142` nodeSelector ` 是节点选择约束的最简单推荐形式。你可以将 ` nodeSelector ` 字段添加到
144143Pod 的规约中设置你希望目标节点所具有的[ 节点标签] ( #built-in-node-labels ) 。
@@ -182,7 +181,7 @@ define. Some of the benefits of affinity and anti-affinity include:
182181The affinity feature consists of two types of affinity:
183182
184183* *Node affinity* functions like the `nodeSelector` field but is more expressive and
185- allows you to specify soft rules.
184+ allows you to specify soft rules.
186185* *Inter-pod affinity/anti-affinity* allows you to constrain Pods against labels
187186 on other Pods.
188187-->
@@ -263,7 +262,7 @@ interpreting the rules. You can use `In`, `NotIn`, `Exists`, `DoesNotExist`,
263262
264263<!--
265264`NotIn` and `DoesNotExist` allow you to define node anti-affinity behavior.
266- Alternatively, you can use [node taints](/docs/concepts/scheduling-eviction/taint-and-toleration/)
265+ Alternatively, you can use [node taints](/docs/concepts/scheduling-eviction/taint-and-toleration/)
267266to repel Pods from specific nodes.
268267-->
269268` NotIn ` 和 ` DoesNotExist ` 可用来实现节点反亲和性行为。
@@ -323,7 +322,7 @@ The final sum is added to the score of other priority functions for the node.
323322Nodes with the highest total score are prioritized when the scheduler makes a
324323scheduling decision for the Pod.
325324
326- For example, consider the following Pod spec:
325+ For example, consider the following Pod spec:
327326-->
328327最终的加和值会添加到该节点的其他优先级函数的评分之上。
329328在调度器为 Pod 作出调度决定时，总分最高的节点的优先级也最高。
@@ -550,7 +549,7 @@ The affinity rule says that the scheduler can only schedule a Pod onto a node if
550549the node is in the same zone as one or more existing Pods with the label
551550` security=S1` . More precisely, the scheduler must place the Pod on a node that has the
552551` topology.kubernetes.io/zone=V` label, as long as there is at least one node in
553- that zone that currently has one or more Pods with the Pod label `security=S1`.
552+ that zone that currently has one or more Pods with the Pod label `security=S1`.
554553-->
555554亲和性规则表示，仅当节点和至少一个已运行且有 `security=S1` 的标签的
556555Pod 处于同一区域时，才可以将该 Pod 调度到节点上。
@@ -615,7 +614,7 @@ affinity/anti-affinity definition appears.
615614-->
616615除了 `labelSelector` 和 `topologyKey`，你也可以指定 `labelSelector`
617616要匹配的命名空间列表，方法是在 `labelSelector` 和 `topologyKey`
618- 所在层同一层次上设置 `namespaces`。
617+ 所在层同一层次上设置 `namespaces`。
619618如果 `namespaces` 被忽略或者为空，则默认为 Pod 亲和性/反亲和性的定义所在的命名空间。
620619
621620<!--
@@ -628,7 +627,7 @@ affinity/anti-affinity definition appears.
628627<!--
629628You can also select matching namespaces using `namespaceSelector`, which is a label query over the set of namespaces.
630629The affinity term is applied to namespaces selected by both `namespaceSelector` and the `namespaces` field.
631- Note that an empty `namespaceSelector` ({}) matches all namespaces, while a null or empty `namespaces` list and
630+ Note that an empty `namespaceSelector` ({}) matches all namespaces, while a null or empty `namespaces` list and
632631null `namespaceSelector` matches the namespace of the Pod where the rule is defined.
633632-->
634633用户也可以使用 `namespaceSelector` 选择匹配的名字空间，`namespaceSelector`
@@ -641,7 +640,7 @@ null `namespaceSelector` matches the namespace of the Pod where the rule is defi
641640# ### More practical use-cases
642641
643642Inter-pod affinity and anti-affinity can be even more useful when they are used with higher
644- level collections such as ReplicaSets, StatefulSets, Deployments, etc. These
643+ level collections such as ReplicaSets, StatefulSets, Deployments, etc. These
645644rules allow you to configure that a set of workloads should
646645be co-located in the same defined topology; for example, preferring to place two related
647646Pods onto the same node.
@@ -664,7 +663,7 @@ affinity and anti-affinity to co-locate the web servers with the cache as much a
664663你可以使用 Pod 间的亲和性和反亲和性来尽可能地将该 Web 服务器与缓存并置。
665664
666665<!--
667- In the following example Deployment for the redis cache, the replicas get the label `app=store`. The
666+ In the following example Deployment for the Redis cache, the replicas get the label `app=store`. The
668667` podAntiAffinity` rule tells the scheduler to avoid placing multiple replicas
669668with the `app=store` label on a single node. This creates each cache in a
670669separate node.
@@ -764,9 +763,9 @@ where each web server is co-located with a cache, on three separate nodes.
764763| *webserver-1* | *webserver-2* | *webserver-3* |
765764| *cache-1* | *cache-2* | *cache-3* |
766765
767- <!--
766+ <!--
768767The overall effect is that each cache instance is likely to be accessed by a single client, that
769- is running on the same node. This approach aims to minimize both skew (imbalanced load) and latency.
768+ is running on the same node. This approach aims to minimize both skew (imbalanced load) and latency.
770769-->
771770总体效果是每个缓存实例都非常可能被在同一个节点上运行的某个客户端访问。
772771这种方法旨在最大限度地减少偏差（负载不平衡）和延迟。
@@ -849,7 +848,7 @@ The above Pod will only run on the node `kube-01`.
849848-->
850849上面的 Pod 只能运行在节点 `kube-01` 之上。
851850
852- <!--
851+ <!--
853852# # Pod topology spread constraints
854853
855854You can use _topology spread constraints_ to control how {{< glossary_tooltip text="Pods" term_id="Pod" >}}
@@ -858,7 +857,7 @@ topology domains that you define. You might do this to improve performance, expe
858857overall utilization.
859858
860859Read [Pod topology spread constraints](/docs/concepts/scheduling-eviction/topology-spread-constraints/)
861- to learn more about how these work.
860+ to learn more about how these work.
862861-->
863862# # Pod 拓扑分布约束 {#pod-topology-spread-constraints}
864863
@@ -877,7 +876,7 @@ to learn more about how these work.
877876* Read the design docs for [node affinity](https://git.k8s.io/design-proposals-archive/scheduling/nodeaffinity.md)
878877 and for [inter-pod affinity/anti-affinity](https://git.k8s.io/design-proposals-archive/scheduling/podaffinity.md).
879878* Learn about how the [topology manager](/docs/tasks/administer-cluster/topology-manager/) takes part in node-level
880- resource allocation decisions.
879+ resource allocation decisions.
881880* Learn how to use [nodeSelector](/docs/tasks/configure-pod-container/assign-pods-nodes/).
882881* Learn how to use [affinity and anti-affinity](/docs/tasks/configure-pod-container/assign-pods-nodes-using-node-affinity/).
883882-->
@@ -888,4 +887,3 @@ to learn more about how these work.
888887* 了解[拓扑管理器](/zh-cn/docs/tasks/administer-cluster/topology-manager/)如何参与节点层面资源分配决定。
889888* 了解如何使用 [nodeSelector](/zh-cn/docs/tasks/configure-pod-container/assign-pods-nodes/)。
890889* 了解如何使用[亲和性和反亲和性](/zh-cn/docs/tasks/configure-pod-container/assign-pods-nodes-using-node-affinity/)。
891-
0 commit comments