In this article, we will take a look at namespaces in Kubernetes, explain what they are and how to create them on the command line and in a configuration file, with some quick reference command examples using kubectl.
What we will cover:
Namespaces in Kubernetes allow the separation and logical organization of groups of resources. Effective use of namespaces can make cluster management more streamlined. They can be created using the command line tool kubectl or using a YAML configuration file.
Namespaces can be used for logical groupings, for example, to denote multiple environments like dev, qa, or prod, or for grouping objects that make up an application, like this blog! Commonly a combination of both the above examples is used. If an application is called blog, then it could be segmented using namespaces blog-dev, blog-qa, and blog-prod.
If a resource is namespaced, that means it cannot be created without a namespace.
Default namespaces in Kubernetes
By default, when you are creating a new Kubernetes cluster using a managed service, the following namespaces are created:
kubectl get namespaces
- default – used for resources that don’t have a specific namespace specified when created
- kube-node-lease – helps determine the availability of the nodes
- kube-public – used for resources that need to be readable by all users
- kube-system – contains the system processes such as DNS, dashboard, logging, and monitoring, among others
Kubernetes cluster vs. namespace
A Kubernetes cluster is spread among multiple compute nodes and contains multiple namespaces in which k8s resources are deployed. Not all resources in kubernetess are namespaced, meaning that some of them can be created globally.
Kubernetes namespace vs. pods
Pods are the most fundamental resources you can use inside your Kubernetes cluster for your workloads. They are namespaced, meaning they are always created inside a Kubernets namespace.
Kubernetes Namespaces are critical for your Kubernetes workloads. They provide a logical separation between applications, environments, and teams.
When to use multiple namespaces?
There are multiple scenarios in which you will need to use multiple namespaces:
- Multi-Tenant Environments
Namespaces provide a way to isolate resources, ensuring that only team members with enough permissions can access them. By using namespaces in multi-tenant environments you can also take advantage of resource quotas and limits, making sure that no teams are overwhelming the cluster with their consumption.
- Environment Separation
Usually, when developing an application, it is best practice to have multiple environments to ensure that all changes are applied safely. With K8s namespaces, you can create one namespace per environment and deploy your application while ensuring that each environment can be managed independently. This is also helpful for credentials management as configmaps and secrets are namespaced resources, and you will most likely have different configurations based on the environment.
- Implementing Role-Based Access Control
Roles and Role Bindings are namespaced resources that allow you to define permissions within the scope of a namespace.
- Testing
Creating a separate namespace and deploying your application will help developers test their application while removing any risk associated with direct production changes.
# namespace.yaml
apiVersion: v1
kind: Namespace
metadata:
name: jacks-blogNow we can create it by running:
# Apply the namespace.yaml file
kubectl apply -f namespace.yml
Using kubectl
To create a Kubernetes namespace imperatively with kubectl, you can simply run this command:
kubectl create namespace my-namespace-imperative
namespace/my-namespace-imperative created
Note that on the command line, if a namespace is not specified for a given kubectl command, then objects from the default namespace are shown.
1. View existing namespaces
To see objects within a specific namespace, the name of the namespace must be specified. Creating objects in the default namespace is considered a bad practice as it makes it harder to implement NetworkPolicies, use RBAC, and segregate objects. (Read more in our Kubernetes best practices article.)
# View existing namespaces
kubectl get namespaces
Here I have a namespace listed called ingress-nginx.
Note that I also see the four Kubernetes namespaces that come out of the box:
- default —The default namespace set by the system.
- kube-system —This namespace is assigned to resources that are created by the Kubernetes system.
- kube-public —This namespace is created by the system and is visible to all users, even users that aren’t authenticated.
- kube-node-lease —This namespace holds lease objects associated with each node. These leases allow the kubelet to send heartbeats so that you can determine node availability.
2. List the pods contained in a namespace
To view the pods within the namespace:
# List the pods contained in a namespace
kubectl get pods --namespace ingress-nginx
# Note the short format for namespace can be used (-n)
kubectl get pods -n ingress-nginx
3. List pods in the default namespace
To view pods in the default namespace (no namespace specified):
# List pods in the default namespace
kubectl get pods
Note that my three pods in the ingress-nginx namespace are not displayed here.
4. Create a new namespace
To create a new namespace:
# Create a new namespace called jacks-blog
kubectl create namespace jacks-blog
The namespace’s name can be created with a maximum length of 63 characters using only alphanumeric characters and hyphens. Names cannot start with a hyphen, and the alpha characters can only be lowercase.
5. Delete a namespace
And to delete a namespace:
# Delete a namespace called jacks-blog
kubectl delete namespace jacks-blog
Read more about the kubectl delete deployment command.
6. See details about a namespace
To see details about a namespace:
# Describe a namespace
kubectl describe namespace nginx-ingress
7. Show resource usage of pods in a namespace
To see the resource usage (CPU / Memory) of pods in a particular namespace:
# Show resource usage of pods in ingress-nginx namespace
kubectl top pod --namespace=ingress-nginx
See also How to Restart Kubernetes Pods With Kubectl.
8. Rename a namespace
While you cannot directly rename your namespace, what you can do is:
- Save the configuration of it to a file.
kubectl get namespace my-namespace-declarative -o yaml > new_namespace.yaml- Modify the name in the yaml file.
- Create the new namespace.
kubectl apply -f new_namespace.yaml
namespace/my-namespace-declarative-modified created
- Repeat the above three steps for all the resources inside your namespace (you will need to modify the namespace parameter and not the name of the resource).
- Delete the old namespace.
kubectl delete namespace my-namespace-declarative
namespace "my-namespace-declarative" deletedNamespaces are an important piece for your K8s cluster. They help with the overall separation, governance, and compliance of your Kubernetes resources.
Here are some of the best practices you should take advantage of:
1. Use meaningful names for your namespaces
Create your namespaces for specific use cases: environment separation, multi-tenant RBAC, testing, and others. By looking at a namespace name, you should be able to identify its purpose easily.
2. Implement resource quotas and limits
On the namespace level, you should ensure that resources have limits for CPU and memory, in order to avoid overloading the cluster. Also, using persistent volume claims for your storage helps in maintaining cluster performance.
3. Implementing role-based access control
Using the least privilege access method is a best practice for every IT tool or product, and K8s makes no difference. You need to create roles and role bindings to your namespaces and ensure that only authorized users have access to them. (To learn more, check out Guide to Kubernetes RBAC)
4. Implement network policies
Network policies are similar to security group / network access control list rules. With them, you can control pod communication and restrict connection between services that don’t need to interact.
5. Take advantage of monitoring and logging
Configure monitoring and logging as part of your cluster and set up alerts based on namespace-specific metrics. This can help in overseeing the cluster’s well-being and detect anomalies, while giving you the option to promptly address these issues.
6. Use multiple namespaces
Don’t create all your resources in a single namespace because, in that way, it will be impossible to implement all the best practices mentioned above. You should find a namespace split that works for your organization (application/environment/team), and put it to the test.
Namespaces are easy to work with in Kubernetes and can be manipulated using the kubectl command line tool, or declared in YAML configuration files. They allow objects to be grouped. Effective use of namespaces can make cluster management more streamlined.
Also, anything that can be run via kubectl can be run within a Spacelift stack. Spacelift helps you manage the complexities and compliance challenges of using Kubernetes. It brings with it a GitOps flow, so your Kubernetes Deployments are synced with your Kubernetes Stacks, and pull requests show you a preview of what they’re planning to change. It also has an extensive selection of policies, which lets you automate compliance checks and build complex multi-stack workflows.
Manage Kubernetes Faster and More Easily
Spacelift allows you to automate, audit, secure, and continuously deliver your infrastructure. It helps overcome common state management issues and adds several must-have features for infrastructure management.
