Limit Range
By default, containers run with unbounded compute resources on a Kubernetes cluster.
LimitRange is a Kubernetes resource that enforces resource constraints (CPU, memory) at the namespace level. It ensures that every pod/container in the namespace has a default resource request and limit. Containers cannot exceed the specified limits. Using this will prevents excessive resource consumption by a single pod/container.
Define LimitRange
Lets create new file limitrange.yaml and put below definition there. This configuration will limit CPU and Memory in container level.
apiVersion: v1
kind: LimitRange
metadata:
name: resource-limits
namespace: limited
spec:
limits:
- type: Container
defaultRequest:
cpu: "100m"
memory: "64Mi"
default:
cpu: "100m"
memory: "64Mi"
max:
cpu: "250m"
memory: "128Mi"defaultRequest: Define the default request for each containers.cpu: "100m": Set container'sresource.request.cputo100mif not explicitly specified.memory: "64Mi": Set container'sresource.request.memoryto64Miif not explicitly specified.
default: Define the default limit for each containers.cpu: "100m": Set container'sresource.limits.cputo100mif not explicitly specified.memory: "64Mi": Set container'sresource.limits.memoryto64Miif not explicitly specified.
max: Define the default request for each containers. The limits for configured in each containers cannot exceed this value.
Apply and Validate
Let's apply the configuration file using kubectl apply command.
➜ kubectl apply -f limitrange.yaml
limitrange/resource-limits createdWe can validate it using kubectl describe limitranges command with specific namespace.
➜ kubectl -n limited describe limitranges
Name: resource-limits
Namespace: limited
Type Resource Min Max
---- -------- --- ---
Container cpu - 250m
Container memory - 128Mi
Default Request Default Limit Max Limit/Request Ratio
--------------- ------------- -----------------------
100m 100m -
64Mi 64Mi -Test LimitRange
To test if our configuration work as expected we can run pod without any resource specified.
➜ kubectl -n limited run nginx --image=nginx
pod/nginx createdLets get the pod details using kubectl describe command.
➜ kubectl -n limited describe pod nginx
Name: nginx
Namespace: limited
Priority: 0
Service Account: default
Node: minikube/192.168.49.2
Start Time: Wed, 05 Feb 2025 17:49:07 +0700
Labels: run=nginx
Annotations: kubernetes.io/limit-ranger: LimitRanger plugin set: cpu, memory request for container nginx; cpu, memory limit for container nginx
Status: Pending
IP:
IPs: <none>
Containers:
nginx:
Container ID:
Image: nginx
Image ID:
Port: <none>
Host Port: <none>
State: Waiting
Reason: ContainerCreating
Ready: False
Restart Count: 0
Limits:
cpu: 100m
memory: 64Mi
Requests:
cpu: 100m
memory: 64Mi
Environment: <none>
...As you can see above if we don't specify the resource.request and resource.limits it will set to default that we just configure.
Now let's try to create a new pod that will exceed the max configuration of our LimitRange. Create new file named limitrange-pod.yaml and put below definition there.
apiVersion: v1
kind: Pod
metadata:
name: over-limit-pod
namespace: limited
spec:
containers:
- name: over-limit-container
image: nginx
resources:
limits:
cpu: "1" # Exceeds the 250m limit
memory: "512Mi" # Exceeds the 128Mi limitLets apply and see what happens.
➜ kubectl apply -f limitrange-pod.yaml
Error from server (Forbidden): error when creating "limitrange-pod.yaml": pods "over-limit-pod" is forbidden: [maximum cpu usage per Container is 250m, but limit is 1, maximum memory usage per Container is 128Mi, but limit is 512Mi]As you can see we got error creating pods "over-limit-pod" is forbidden because the CPU and Memory limits exceed the LimitRange.
References
Last updated