17 Kubernetes interview questions

Kubernetes is a system that is used to automate deployment, scaling, and database management. Kubernetes is an open-source system that was originally designed by Google but is now managed by the Cloud Native Computing Foundation.

Kubernetes works with a large number of container tools and runs containers in a cluster.

Some of the main features of Kubernetes are:

• Automated Scheduling
• Self Healing Capabilities
• Automated rollouts and rollback
• Load Balancing
• Horizontal Scaling

Integration of multiple services that allow you to automate processes is called Orchestration. Orchestration allows you to schedule work and synchronizes information periodically.

Let’s assume we are using multiple containers and each performs a different task. Orchestration helps these containers communicate and execute tasks in an automated and efficient manner.

Nodes make up the primary components of Kubernetes architecture. Nodes in Kubernetes are the smallest unit of hardware and represent a machine in a cluster.

Nodes are managed by a master component that controls and schedule the pods. Some of these node services include Kubelet, Kube-proxy, etc.

Pods in Kubernetes are high level-structures that are used to wrap containers. Pods allow the containers to communicate with other containers in a pod. They also facilitate sharing of local networks and resources. A Pod always runs on a node.

There are two different types of pods:

• Single Container Pod
• Multi Container Pods

Daemon sets are used to perform host layer attributes that help monitor the network. Since it is applied to the host, it only needs to be executed once.

Namespaces are used to divide cluster resources between users. Because of this behavior, it is commonly used in environments where there are multiple users.

Namespaces are also used to organize the application into groups, allowing them to be monitored more easily.

The initial namespaces that Kubernetes starts with are:

• Default
• Kube - system
• Kube - public

The various services provided within Kubernetes are:

• Cluster IP
• Load Balancing
• External name creation
• Node Port

Load balancing helps distribute the traffic onto multiple services. It helps divide the load on the backend server and also ensures that the application is available to all the users.

Load balancing in Kubernetes is done using the following method. The entire load is directed towards a single IP address. The subsequent load is then redirected to other servers using an algorithm known as round-robin scheduling.

There are two types of load balancers:

• Internal load balancer
• External load balancer

The various tools that are available and used to monitor a container are:

• Heapster
• cAdvisor
• Prometheus
• Grafana

Google Kubernetes Engine or GKE, is an open-source system used to manage containers and clusters. However, this Kubernetes engine supports only those clusters that run on google’s public cloud service.

The various types of Kubernetes Volume are:

• GCE persistent disk
• Flocker
• HostPath
• NFS
• ISCSI
• rbd
• PersistentVolumeClaim
• downwardAPI

Stateful sets are workload APIs used to manage stateful applications. It can also be used to scale sets of pods and also to manage their deployments.

The state information and other related data of the pods are all stored on disk storage.

Kubectl is used to control Kubernetes clusters. It uses a command-line interface to communicate and manage the cluster.

A few of the commonly used commands are:

• kubectl annotate
• kubectl cluster-info
• kubectl attach
• kubectl apply
• kubectl config
• kubectl autoscale
• kubectl config current-context
• kubectl config set

Labels are a collection of keys that contain values. The key values are connected to pods, replication controllers, and associated services.

Generally, labels are added to some object during its creation time. They can be modified by the users at run time.

The ways to provide API-Security on Kubernetes are:

• Use the correct auth mode with API server authentication mode= Node.
• Make kubeless to protect its API via authorization-mode=Webhook.
• Ensure the kube-dashboard uses a restrictive RBAC (Role-Based Access Control) policy