资源配额(ResourceQuota) & 资源限制(LimitRange)

资源配额 ResourceQuota

资源配额 ResourceQuota:限制命名空间总容量。

当多个团队、多个用户共享使用K8s集群时,会出现不均匀资源使用,默认情况下先到先得,这时可以通过ResourceQuota来对命名空间资源使用总量做限制,从而解决这个问题。

使用流程:k8s管理员为每个命名空间创建一个或多个ResourceQuota对象,定义资源使用总量,K8s会跟踪命名空间资源使用情况,当超过定义的资源配额会返回拒绝

image-20230524202104161

💘 实战:资源配额 ResourceQuota-2023.5.25(测试成功)

image-20230524212414355

  • 实验环境
实验环境:
1、win10,vmwrokstation虚机;
2、k8s集群:3台centos7.6 1810虚机,1个master节点,2个node节点
   k8s version:v1.20.0
   docker://20.10.7
  • 实验软件

链接:https://pan.baidu.com/s/1-NzdrpktfaUOAl6QO_hqsA?pwd=0820
提取码:0820
2023.5.25-ResourceQuota-code

image-20230525070729388

1.计算资源配额

image-20230525061309728

  • 自己虚机目前配置为2c,2g

[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-mFzQVe8l-1685312821737)(https://bucket-hg.oss-cn-shanghai.aliyuncs.com/img/image-20230524213037637.png)]

为1master,2个node节点。

  • 创建测试命名空间test
[root@k8s-master1 ~]#kubectl create ns test
namespace/test created
  • 创建ResourceQuota资源
[root@k8s-master1 ~]#mkdir ResourceQuota
[root@k8s-master1 ~]#cd ResourceQuota/
[root@k8s-master1 ResourceQuota]#vim compute-resources.yaml
apiVersion: v1
kind: ResourceQuota
metadata:
  name: compute-resources
  namespace: test
spec:
  hard:
    requests.cpu: "1"
    requests.memory: 1Gi
    limits.cpu: "2"
    limits.memory: 2Gi
    
#部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f compute-resources.yaml 
resourcequota/compute-resources configured

#查看当前配置的ResourceQuota
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE     REQUEST                                     LIMIT
compute-resources   2m37s   requests.cpu: 0/1, requests.memory: 0/1Gi   limits.cpu: 0/2, limits.memory: 0/2Gi
  • 部署一个pod应用
[root@k8s-master1 ResourceQuota]#kubectl run web --image=nginx --dry-run=client -oyaml > pod.yaml
[root@k8s-master1 ResourceQuota]#vim pod.yaml
#删除没用的配置,并配置上resources
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 0.5
        memory: 0.5Gi
      limits:
        cpu: 1
        memory: 1Gi        
        
#部署pod
[root@k8s-master1 ResourceQuota]#kubectl apply -f pod.yaml 
Error from server (Forbidden): error when creating "pod.yaml": pods "web" is forbidden: failed quota: compute-resources: must specify limits.cpu,limits.memory
#注意:在部署pod时会看到报错,提示"pods "web" is forbidden: failed quota: compute-resources: must specify limits.cpu,limits.memory",因为test命名空间配置了ResourceQuota,pod里只配置requests会报错;

#测试:如果不配置resource,看会否会报错?
cat pod.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    #resources:
     # requests:
      #  cpu: 0.5
      #  memory: 0.5Gi
[root@k8s-master1 ResourceQuota]#kubectl apply -f pod.yaml 
Error from server (Forbidden): error when creating "pod.yaml": pods "web" is forbidden: failed quota: compute-resources: must specify limits.cpu,limits.memory,requests.cpu,requests.memory
[root@k8s-master1 ResourceQuota]#kubectl  get po -ntest
No resources found in test namespace.
#现象:也是会报错的!!!

#结论:只要是配置了ResourceQuota的命名空间,pod里必须要配置上limits.cpu,limits.memory,requests.cpu,requests.memory,否则`会返回拒绝`,无法成功创建资源的。

#重新配置pod:补加limits配置
[root@k8s-master1 ResourceQuota]#vim pod.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 0.5
        memory: 0.5Gi
      limits:
        cpu: 1
        memory: 1Gi 
        
#重新部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f pod.yaml 
pod/web created
  • 查看
#查看:
[root@k8s-master1 ResourceQuota]#kubectl get po -ntest
NAME   READY   STATUS    RESTARTS   AGE
web    1/1     Running   0          26s
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE   REQUEST                                            LIMIT
compute-resources   8h    requests.cpu: 500m/1, requests.memory: 512Mi/1Gi   limits.cpu: 1/2, limits.memory: 1Gi/2Gi
#可以看到,此时ResourceQuota下可以清楚地看到requests.cpu,requests.memory,limits.cpu,limits.memory的当前使用量/总使用量。
  • 这里测试下,若继续在test命名空间下新建pod,如果pod里cpu或者memory的requests值之和超过ResourceQuota里定义的,预计会报错。当然,pod里cpu或者memory的rlimits值之和超过ResourceQuota里定义的,同理也会报错。接下来,我们测试下看看。

测试:如果pod里cpu或者memory的requests值之和超过ResourceQuota里定义的,预计会报错。

[root@k8s-master1 ResourceQuota]#cp pod.yaml pod1.yaml
[root@k8s-master1 ResourceQuota]#vim pod1.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web2
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 0.6
        memory: 0.5Gi
      limits:
        cpu: 1
        memory: 1Gi

#部署,并观察现象:
[root@k8s-master1 ResourceQuota]#kubectl apply -f pod1.yaml 
Error from server (Forbidden): error when creating "pod1.yaml": pods "web2" is forbidden: exceeded quota: compute-resources, requested: requests.cpu=600m, used: requests.cpu=500m, limited: requests.cpu=1
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE   REQUEST                                            LIMIT
compute-resources   8h    requests.cpu: 500m/1, requests.memory: 512Mi/1Gi   limits.cpu: 1/2, limits.memory: 1Gi/2Gi

结论:如果pod里cpu或者memory的requests值之和超过ResourceQuota里定义的,创建新的pod会报错。

测试:如果pod里cpu或者memory的linits值之和超过ResourceQuota里定义的,预计会报错。

[root@k8s-master1 ResourceQuota]#cp pod.yaml pod2.yaml
[root@k8s-master1 ResourceQuota]#vim pod2.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web3
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 0.5
        memory: 0.5Gi
      limits:
        cpu: 1.1
        memory: 1Gi


#部署,并观察现象:
[root@k8s-master1 ResourceQuota]#kubectl apply -f pod2.yaml 
Error from server (Forbidden): error when creating "pod2.yaml": pods "web3" is forbidden: exceeded quota: compute-resources, requested: limits.cpu=1100m, used: limits.cpu=1, limited: limits.cpu=2
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE   REQUEST                                            LIMIT
compute-resources   8h    requests.cpu: 500m/1, requests.memory: 512Mi/1Gi   limits.cpu: 1/2, limits.memory: 1Gi/2Gi

结论:如果pod里cpu或者memory的limits值之和超过ResourceQuota里定义的,创建新的pod会报错。

因此:

如果某个命名空间下配置了ResourceQuota,pod里必须要配置上limits.cpu,limits.memory,requests.cpu,requests.memory,否则会返回拒绝,无法成功创建资源的。

另外,如果pod里cpu或者memory的requests&limits值之和超过ResourceQuota里定义的requests&limits,则会返回拒绝,无法成功创建资源的。

[root@k8s-master1 ResourceQuota]#cp pod.yaml pod3.yaml
[root@k8s-master1 ResourceQuota]#vim pod3.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web3
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 0.5
        memory: 0.5Gi
      limits:
        cpu: 1
        memory: 1Gi

#部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f pod3.yaml
pod/web3 created

#查看:
[root@k8s-master1 ResourceQuota]#kubectl get po -ntest
NAME   READY   STATUS    RESTARTS   AGE
web    1/1     Running   0          16m
web3   1/1     Running   0          27s
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE   REQUEST                                       LIMIT
compute-resources   8h    requests.cpu: 1/1, requests.memory: 1Gi/1Gi   limits.cpu: 2/2, limits.memory: 2Gi/2Gi

测试结束。😘

2.存储资源配额

image-20230525061406558

  • 部署ResourceQuota
[root@k8s-master1 ResourceQuota]#vim storage-resources.yaml
apiVersion: v1
kind: ResourceQuota
metadata:
  name: storage-resources
  namespace: test
spec:
  hard:
    requests.storage: "10G"  
    
#部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f storage-resources.yaml 
resourcequota/storage-resources created

#查看
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE   REQUEST                                       LIMIT
compute-resources   8h    requests.cpu: 1/1, requests.memory: 1Gi/1Gi   limits.cpu: 2/2, limits.memory: 2Gi/2Gi
storage-resources   8s    requests.storage: 0/10G
  • 创建pvc测试
[root@k8s-master1 ResourceQuota]#vim pvc.yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pvc
  namespace: test
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 8Gi
      
#部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f pvc.yaml 
persistentvolumeclaim/pvc created

#查看:
[root@k8s-master1 ResourceQuota]#kubectl get pvc -ntest
NAME   STATUS    VOLUME   CAPACITY   ACCESS MODES   STORAGECLASS   AGE
pvc    Pending     #这个pending不影响实验,这里是没有pv才导致pvc处于Pending状态    
#部署成功后,可以看到ResourceQuota requests.storage这里已经发生了变化
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE    REQUEST                                       LIMIT
compute-resources   8h     requests.cpu: 1/1, requests.memory: 1Gi/1Gi   limits.cpu: 2/2, limits.memory: 2Gi/2Gi
storage-resources   117s   requests.storage: 8Gi/10G
  • 我们继续在创建一个pvc,此时如果requests.storage之和超过ResourceQuota里定义的话,那么预计会报错的
[root@k8s-master1 ResourceQuota]#cp pvc.yaml pvc1.yaml
[root@k8s-master1 ResourceQuota]#vim pvc1.yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pvc1
  namespace: test
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 2.1Gi
      
#部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f pvc1.yaml 
Error from server (Forbidden): error when creating "pvc1.yaml": persistentvolumeclaims "pvc1" is forbidden: exceeded quota: storage-resources, requested: requests.storage=2254857831, used: requests.storage=8Gi, limited: requests.storage=10G
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE     REQUEST                                       LIMIT
compute-resources   8h      requests.cpu: 1/1, requests.memory: 1Gi/1Gi   limits.cpu: 2/2, limits.memory: 2Gi/2Gi
storage-resources   6m29s   requests.storage: 8Gi/10G
#可以看到,此时报错了,意料之中。

#我们重新部署一个pvc看看:
[root@k8s-master1 ResourceQuota]#cp pvc.yaml pvc2.yaml
[root@k8s-master1 ResourceQuota]#vim pvc2.yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pvc2
  namespace: test
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 2Gi
#部署:注意,存储这里不能使用满!!!
[root@k8s-master1 ResourceQuota]#kubectl apply -f pvc2.yaml 
Error from server (Forbidden): error when creating "pvc2.yaml": persistentvolumeclaims "pvc2" is forbidden: exceeded quota: storage-resources, requested: requests.storage=2Gi, used: requests.storage=8Gi, limited: requests.storage=10G
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE     REQUEST                                       LIMIT
compute-resources   8h      requests.cpu: 1/1, requests.memory: 1Gi/1Gi   limits.cpu: 2/2, limits.memory: 2Gi/2Gi
storage-resources   6m29s   requests.storage: 8Gi/10G

#我们再次创建下pvc3.yaml看下
[root@k8s-master1 ResourceQuota]#cp pvc.yaml pvc3.yaml
[root@k8s-master1 ResourceQuota]#vim pvc3.yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pvc3
  namespace: test
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 1Gi

#部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f pvc3.yaml 
persistentvolumeclaim/pvc3 created

#查看:符合预期,可以正常部署pvc。
[root@k8s-master1 ResourceQuota]#kubectl get pvc -ntest
NAME   STATUS    VOLUME   CAPACITY   ACCESS MODES   STORAGECLASS   AGE
pvc    Pending                                                     9m27s
pvc3   Pending                                                     9s
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE   REQUEST                                       LIMIT
compute-resources   8h    requests.cpu: 1/1, requests.memory: 1Gi/1Gi   limits.cpu: 2/2, limits.memory: 2Gi/2Gi
storage-resources   11m   requests.storage: 9Gi/10G
[root@k8s-master1 ResourceQuota]#

测试结束。😘

3.对象数量配额

image-20230525063637024

  • 我们来看下当前环境
[root@k8s-master1 ResourceQuota]#kubectl get po -ntest
NAME   READY   STATUS    RESTARTS   AGE
web    1/1     Running   0          41m
web3   1/1     Running   0          25m
  • 部署ResourceQuota
[root@k8s-master1 ResourceQuota]#vim object-counts.yaml
apiVersion: v1
kind: ResourceQuota
metadata:
  name: object-counts
  namespace: test
spec:
  hard:
    pods: "4"
    count/deployments.apps: "3"
    count/services: "3"
    
#部署:
[root@k8s-master1 ResourceQuota]#kubectl apply -f object-counts.yaml 
resourcequota/object-counts created

#查看:
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE   REQUEST                                                       LIMIT
compute-resources   9h    requests.cpu: 1/1, requests.memory: 1Gi/1Gi                   limits.cpu: 2/2, limits.memory: 2Gi/2Gi
object-counts       15s   count/deployments.apps: 0/3, count/services: 0/3, pods: 2/4
storage-resources   16m   requests.storage: 9Gi/10G
[root@k8s-master1 ResourceQuota]#
  • 测试
#此时已经存在2个pod了,ResourceQuota里限制的pod最大数量为4,那我们接下来创建下测试pod看下
#但此时为了测试方方便,我删除下前面的compute-resources.yaml,不然创建pod会报错的
[root@k8s-master1 ResourceQuota]#kubectl delete -f compute-resources.yaml 
resourcequota "compute-resources" deleted
[root@k8s-master1 ResourceQuota]#kubectl get quota -ntest
NAME                AGE     REQUEST                                                       LIMIT
object-counts       3m23s   count/deployments.apps: 0/3, count/services: 0/3, pods: 2/4
storage-resources   19m     requests.storage: 9Gi/10G

#创建3个测试pod看下
[root@k8s-master1 ResourceQuota]#kubectl get po -ntest
NAME   READY   STATUS    RESTARTS   AGE
web    1/1     Running   0          47m
web3   1/1     Running   0          32m
[root@k8s-master1 ResourceQuota]#kubectl run web4 --image=nginx -ntest
pod/web4 created
[root@k8s-master1 ResourceQuota]#kubectl run web5 --image=nginx -ntest
Error from server (Forbidden): pods "web5" is forbidden: exceeded quota: object-counts, requested: pods=1, used: pods=4, limited: pods=4
#可以看到,这里报错了。

测试结束。😘

汇总

  • 注意事项
  1. 如果某个命名空间下配置了ResourceQuota,pod里必须要配置上limits.cpu,limits.memory,requests.cpu,requests.memory,否则会返回拒绝,无法成功创建资源的。
  2. 如果pod里cpu或者memory的requests&limits值之和超过ResourceQuota里定义的requests&limits,则会返回拒绝,无法成功创建资源的。(需要注意:实际创建的request和limits pod之和是可以等于这个ResourceQuota定义的数值的,但是存储资源配额:requests.storage、对象数量配额是不能超过(必须小于)ResourceQuota定义的数值,否则会报错的。)
  • 这些字段是可以写在一起的

image-20230525064623650

资源限制 LimitRange

资源限制 LimitRange:限制容器的最大最小。

默认情况下,K8s集群上的容器对计算资源没有任何限制,可能会导致个别容器资源过大导致影响其他容器正常工作,这时可以使用LimitRange定义容器默认CPU和内存请求值或者最大上限

LimitRange限制维度:

• 限制容器配置requests.cpu/memory,limits.cpu/memory的最小、最大值

• 限制容器配置requests.cpu/memory,limits.cpu/memory的默认值

• 限制PVC配置requests.storage的最小、最大值

💘 实战:资源限制 LimitRange-2023.5.25(测试成功)

image-20230524212613262

  • 实验环境
实验环境:
1、win10,vmwrokstation虚机;
2、k8s集群:3台centos7.6 1810虚机,1个master节点,2个node节点
   k8s version:v1.20.0
   docker://20.10.7
  • 实验软件

链接:https://pan.baidu.com/s/1dTuFjqToJaCiHHvtYH9xiw?pwd=0820
提取码:0820

2023.5.25-LimitRange-cdoe

image-20230525122821896

  • 为了保持环境的纯洁性,这里删除上面的ResourceQuota配置
[root@k8s-master1 ~]#cd ResourceQuota/
[root@k8s-master1 ResourceQuota]#ls
compute-resources.yaml  object-counts.yaml  pod1.yaml  pod2.yaml  pod3.yaml  pod.yaml  pvc1.yaml  pvc2.yaml  pvc3.yaml  pvc.yaml  storage-resources.yaml
[root@k8s-master1 ResourceQuota]#kubectl delete -f .
resourcequota "object-counts" deleted
pod "web" deleted
pod "web3" deleted
pod "web3" deleted
persistentvolumeclaim "pvc" deleted
persistentvolumeclaim "pvc3" deleted
resourcequota "storage-resources" deleted
Error from server (NotFound): error when deleting "compute-resources.yaml": resourcequotas "compute-resources" not found
Error from server (NotFound): error when deleting "pod1.yaml": pods "web2" not found
Error from server (NotFound): error when deleting "pvc1.yaml": persistentvolumeclaims "pvc1" not found
Error from server (NotFound): error when deleting "pvc2.yaml": persistentvolumeclaims "pvc2" not found
[root@k8s-master1 ResourceQuota]#

1.计算资源最大、最小限制

image-20230525071051206

  • 结论:

如果命名空间里配置了LimitRange,那么后续创建的pod里容器的request值不能小于LimitRange里定义的min(request的最小值),limits值不能大于LimitRange里定义的max(limits的最大值),否则会报错的。

接下来,我们验证下上面的结论。

  • 部署LimitRange
[root@k8s-master1 ~]#mkdir LimitRange
[root@k8s-master1 ~]#cd LimitRange/
[root@k8s-master1 LimitRange]#vim cpu-memory-min-max.yaml
apiVersion: v1
kind: LimitRange
metadata:
  name: cpu-memory-min-max
  namespace: test
spec:
  limits:
  - max: # 容器能设置limit的最大值
      cpu: 1
      memory: 1Gi
    min: # 容器能设置request的最小值
      cpu: 200m 
      memory: 200Mi
    type: Container
 
#部署
[root@k8s-master1 LimitRange]#kubectl apply -f cpu-memory-min-max.yaml 
limitrange/cpu-memory-min-max created

#查看
[root@k8s-master1 LimitRange]#kubectl get limits -n test
NAME                 CREATED AT
cpu-memory-min-max   2023-05-24T23:26:13Z
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   memory    200Mi  1Gi  1Gi              1Gi            -
Container   cpu       200m   1    1                1              -
#注意:这里是有request和limit默认值的
  • 我们来创建一个小于request最小值,创建一个大于limit最大值的pod看下情况

创建一个小于request最小值:

[root@k8s-master1 LimitRange]#vim pod.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 100m
        memory: 200Mi
      limits:
        cpu: 1
        memory: 1Gi
        
#部署:(符合预期效果)
[root@k8s-master1 LimitRange]#kubectl apply -f pod.yaml
Error from server (Forbidden): error when creating "pod.yaml": pods "web" is forbidden: minimum cpu usage per Container is 200m, but request is 100m
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   cpu       200m   1    1                1              -
Container   memory    200Mi  1Gi  1Gi              1Gi            -

创建一个大于limit最大值:

[root@k8s-master1 LimitRange]#cp pod.yaml pod1.yaml
[root@k8s-master1 LimitRange]#vim pod1.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 200m
        memory: 200Mi
      limits:
        cpu: 1.1
        memory: 1Gi
        
#部署:(符合预期效果)
[root@k8s-master1 LimitRange]#kubectl apply -f pod1.yaml 
Error from server (Forbidden): error when creating "pod1.yaml": pods "web" is forbidden: maximum cpu usage per Container is 1, but limit is 1100m
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   memory    200Mi  1Gi  1Gi              1Gi            -
Container   cpu       200m   1    1                1              -

创建一个合适的pod:

[root@k8s-master1 LimitRange]#cp pod.yaml pod2.yaml
[root@k8s-master1 LimitRange]#vim pod2.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web
  namespace: test
spec:
  containers:
  - image: nginx
    name: web
    resources:
      requests:
        cpu: 250m
        memory: 200Mi
      limits:
        cpu: 0.9
        memory: 1Gi
        
#部署:
[root@k8s-master1 LimitRange]#kubectl apply -f pod2.yaml 
pod/web created

#查看:
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   cpu       200m   1    1                1              -
Container   memory    200Mi  1Gi  1Gi              1Gi            -
[root@k8s-master1 LimitRange]#kubectl get po -ntest
NAME   READY   STATUS    RESTARTS   AGE
web    1/1     Running   0          20s
web4   1/1     Running   0          52m

测试结束。😘

2.计算资源默认值限制

[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-q2v0vH4B-1685312821739)(https://bucket-hg.oss-cn-shanghai.aliyuncs.com/img/image-20230525071058992.png)]

  • 结论:

只要给某个命名空间配置了LimitRange,如果你的pod里不配置request、limit,也是可以创建成功pod的,则k8s会默认分配一个request、limit值的。

  • 我们来看下default命名空间下的pod是否有默认request、limit值
root@k8s-master1 LimitRange]#kubectl get limits 
No resources found in default namespace.
[root@k8s-master1 LimitRange]#kubectl get po
NAME       READY   STATUS    RESTARTS   AGE
busybox    1/1     Running   6          3d10h
busybox2   1/1     Running   6          3d10h
py-k8s     1/1     Running   1          24h
[root@k8s-master1 LimitRange]#kubectl describe po busybox 
Name:         busybox
Namespace:    default
Priority:     0
Node:         k8s-node2/172.29.9.33
Start Time:   Sun, 21 May 2023 20:44:21 +0800
Labels:       run=busybox
Annotations:  cni.projectcalico.org/podIP: 10.244.169.162/32
……

#可以看到,default命名空间下的pod是没有默认request、limit值的,因此其pod可以用尽宿主机的资源
  • 我们看下test命令空间下,创建一个新pod,是否会有默认request、limit值
[root@k8s-master1 LimitRange]#kubectl get limits -ntest
NAME                 CREATED AT
cpu-memory-min-max   2023-05-24T23:26:13Z
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   cpu       200m   1    1                1              -
Container   memory    200Mi  1Gi  1Gi              1Gi            -
[root@k8s-master1 LimitRange]#kubectl run web520 --image=nginx -ntest
pod/web520 created
[root@k8s-master1 LimitRange]#kubectl describe pod web520 -ntest
Name:         web520
Namespace:    test
Priority:     0
Node:         k8s-node2/172.29.9.33
Start Time:   Thu, 25 May 2023 07:44:16 +0800
Labels:       run=web520
Annotations:  cni.projectcalico.org/podIP: 10.244.169.167/32
              cni.projectcalico.org/podIPs: 10.244.169.167/32
              kubernetes.io/limit-ranger: LimitRanger plugin set: cpu, memory request for container web520; cpu, memory limit for container web520
Status:       Running
IP:           10.244.169.167
IPs:
  IP:  10.244.169.167
Containers:
  web520:
    Container ID:   docker://6cf523d8b462fdfcb44348e4af5247f8c35e5cb0cf4c7e5b0dadaeea76aa8bec
    Image:          nginx
    Image ID:       docker-pullable://nginx@sha256:af296b188c7b7df99ba960ca614439c99cb7cf252ed7bbc23e90cfda59092305
    Port:           <none>
    Host Port:      <none>
    State:          Running
      Started:      Thu, 25 May 2023 07:44:26 +0800
    Ready:          True
    Restart Count:  0
    Limits:
      cpu:     1
      memory:  1Gi
    Requests:
      cpu:        1
      memory:     1Gi
    Environment:  <none>
    
#可以看到,是有默认值的    
  • 接下来,我们更改下这个默认值
[root@k8s-master1 LimitRange]#vim default-cpu-memory-min-max.yaml
apiVersion: v1
kind: LimitRange
metadata:
  name: default-cpu-memory-min-max 
  namespace: test
spec:
  limits:
  - default:
      cpu: 500m
      memory: 500Mi
    defaultRequest:
      cpu: 300m
      memory: 300Mi
    type: Container
    
#部署:
[root@k8s-master1 LimitRange]#kubectl apply -f default-cpu-memory-min-max.yaml 
limitrange/default-cpu-memory-min-max created
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   cpu       200m   1    1                1              -
Container   memory    200Mi  1Gi  1Gi              1Gi            -


Name:       default-cpu-memory-min-max
Namespace:  test
Type        Resource  Min  Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---  ---  ---------------  -------------  -----------------------
Container   cpu       -    -    300m             500m           -
Container   memory    -    -    300Mi            500Mi          -
#可以看到,此时的默认值改变了

#我们再创建一个pod看下现象
[root@k8s-master1 LimitRange]#kubectl run web1314 --image=nginx -ntest
pod/web1314 created
[root@k8s-master1 LimitRange]#kubectl describe pod web1314 -ntest
Name:         web1314
Namespace:    test
Priority:     0
Node:         k8s-node1/172.29.9.32
Start Time:   Thu, 25 May 2023 07:51:28 +0800
Labels:       run=web1314
Annotations:  cni.projectcalico.org/podIP: 10.244.36.101/32
              cni.projectcalico.org/podIPs: 10.244.36.101/32
              kubernetes.io/limit-ranger: LimitRanger plugin set: cpu, memory request for container web1314; cpu, memory limit for container web1314
Status:       Running
IP:           10.244.36.101
IPs:
  IP:  10.244.36.101
Containers:
  web1314:
    Container ID:   docker://9cd7d7d0ebf7adb9e71b33749c26dd9e9c0149c3c1e427b50dd733bd3989d75b
    Image:          nginx
    Image ID:       docker-pullable://nginx@sha256:af296b188c7b7df99ba960ca614439c99cb7cf252ed7bbc23e90cfda59092305
    Port:           <none>
    Host Port:      <none>
    State:          Running
      Started:      Thu, 25 May 2023 07:51:30 +0800
    Ready:          True
    Restart Count:  0
    Limits:
      cpu:     1
      memory:  1Gi
    Requests:
      cpu:        1
      memory:     1Gi
    Environment:  <none>
    
#注意:测试出现了问题
#在test命名空间有2个LimitRange资源,其default配置有冲突,但根据测试现象看,默认匹配第一条规则。
#此时,我们直接在第一条规则上直接更改默认值,我们看下现象:
[root@k8s-master1 LimitRange]#vim cpu-memory-min-max.yaml
apiVersion: v1
kind: LimitRange
metadata:
  name: cpu-memory-min-max
  namespace: test
spec:
  limits:
  - max:
      cpu: 1
      memory: 1Gi
    min:
      cpu: 200m
      memory: 200Mi
    default:
      cpu: 600m
      memory: 600Mi
    defaultRequest:
      cpu: 400m
      memory: 400Mi
    type: Container

#部署并查看:
[root@k8s-master1 LimitRange]#kubectl apply -f cpu-memory-min-max.yaml 
limitrange/cpu-memory-min-max configured
root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   cpu       200m   1    400m             600m           -
Container   memory    200Mi  1Gi  400Mi            600Mi          -


Name:       default-cpu-memory-min-max
Namespace:  test
Type        Resource  Min  Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---  ---  ---------------  -------------  -----------------------
Container   cpu       -    -    300m             500m           -
Container   memory    -    -    300Mi            500Mi          -

#再次部署一个新pod
[root@k8s-master1 LimitRange]#kubectl run web1315 --image=nginx -ntest
pod/web1315 created
[root@k8s-master1 LimitRange]#kubectl describe pod web1315  -ntest
Name:         web1315
Namespace:    test
Priority:     0
Node:         k8s-node2/172.29.9.33
Start Time:   Thu, 25 May 2023 07:57:39 +0800
Labels:       run=web1315
Annotations:  cni.projectcalico.org/podIP: 10.244.169.168/32
              cni.projectcalico.org/podIPs: 10.244.169.168/32
              kubernetes.io/limit-ranger: LimitRanger plugin set: cpu, memory request for container web1315; cpu, memory limit for container web1315
Status:       Running
IP:           10.244.169.168
IPs:
  IP:  10.244.169.168
Containers:
  web1315:
    Container ID:   docker://a9e19c184eaa9513078b46d8d2d8ce5dd0a09dca00772052345d59040c57346e
    Image:          nginx
    Image ID:       docker-pullable://nginx@sha256:af296b188c7b7df99ba960ca614439c99cb7cf252ed7bbc23e90cfda59092305
    Port:           <none>
    Host Port:      <none>
    State:          Running
      Started:      Thu, 25 May 2023 07:57:42 +0800
    Ready:          True
    Restart Count:  0
    Limits:
      cpu:     600m
      memory:  600Mi
    Requests:
      cpu:        400m
      memory:     400Mi
    Environment:  <none>
    Mounts:
#此时,新建pod里的默认值被改变过来了,符合预期效果。    

测试结束。😘

3.存储资源最大、最小限制

image-20230525071106317

  • 和前面的计算资源一样,接下来进行测试
[root@k8s-master1 LimitRange]#vim storage-min-max.yaml
apiVersion: v1
kind: LimitRange
metadata:
  name: storage-min-max
  namespace: test
spec:
  limits:
    - type: PersistentVolumeClaim
      max:
        storage: 10Gi
      min:
        storage: 1Gi    

#部署:
[root@k8s-master1 LimitRange]#kubectl apply -f storage-min-max.yaml 
limitrange/storage-min-max created

#创建一个pvc
[root@k8s-master1 LimitRange]#vim pvc.yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pvc-storage-test
  namespace: test
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 11Gi
      
#部署pvc:
[root@k8s-master1 LimitRange]#kubectl apply -f pvc.yaml 
Error from server (Forbidden): error when creating "pvc.yaml": persistentvolumeclaims "pvc-storage-test" is forbidden: maximum storage usage per PersistentVolumeClaim is 10Gi, but request is 11Gi
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   cpu       200m   1    400m             600m           -
Container   memory    200Mi  1Gi  400Mi            600Mi          -


Name:       default-cpu-memory-min-max
Namespace:  test
Type        Resource  Min  Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---  ---  ---------------  -------------  -----------------------
Container   cpu       -    -    300m             500m           -
Container   memory    -    -    300Mi            500Mi          -


Name:                  storage-min-max
Namespace:             test
Type                   Resource  Min  Max   Default Request  Default Limit  Max Limit/Request Ratio
----                   --------  ---  ---   ---------------  -------------  -----------------------
PersistentVolumeClaim  storage   1Gi  10Gi  -                -              -
#可以看到,自己申请的存储为11Gi,但是LimitRange里max可申请存储为10Gi,因此报错。

#重新创建一个pvc
[root@k8s-master1 LimitRange]#cp pvc.yaml pvc1.yaml
[root@k8s-master1 LimitRange]#vim pvc1.yaml 
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pvc-storage-test
  namespace: test
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 8Gi 
      
#部署:
[root@k8s-master1 LimitRange]#kubectl apply -f pvc1.yaml 
persistentvolumeclaim/pvc-storage-test created
[root@k8s-master1 LimitRange]#kubectl get pvc -ntest
NAME               STATUS    VOLUME   CAPACITY   ACCESS MODES   STORAGECLASS   AGE
pvc-storage-test   Pending                                                     30s
[root@k8s-master1 LimitRange]#kubectl describe limits -ntest
Name:       cpu-memory-min-max
Namespace:  test
Type        Resource  Min    Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---    ---  ---------------  -------------  -----------------------
Container   memory    200Mi  1Gi  400Mi            600Mi          -
Container   cpu       200m   1    400m             600m           -


Name:       default-cpu-memory-min-max
Namespace:  test
Type        Resource  Min  Max  Default Request  Default Limit  Max Limit/Request Ratio
----        --------  ---  ---  ---------------  -------------  -----------------------
Container   memory    -    -    300Mi            500Mi          -
Container   cpu       -    -    300m             500m           -


Name:                  storage-min-max
Namespace:             test
Type                   Resource  Min  Max   Default Request  Default Limit  Max Limit/Request Ratio
----                   --------  ---  ---   ---------------  -------------  -----------------------
PersistentVolumeClaim  storage   1Gi  10Gi  -                -              -
#部署成功,符合预期。

测试结束。😘

ibe limits -ntest
Name: cpu-memory-min-max
Namespace: test
Type Resource Min Max Default Request Default Limit Max Limit/Request Ratio


Container memory 200Mi 1Gi 400Mi 600Mi -
Container cpu 200m 1 400m 600m -

Name: default-cpu-memory-min-max
Namespace: test
Type Resource Min Max Default Request Default Limit Max Limit/Request Ratio


Container memory - - 300Mi 500Mi -
Container cpu - - 300m 500m -

Name: storage-min-max
Namespace: test
Type Resource Min Max Default Request Default Limit Max Limit/Request Ratio


PersistentVolumeClaim storage 1Gi 10Gi - - -
#部署成功,符合预期。


测试结束。😘



  • 0
    点赞
  • 1
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值