Concepts

Question 1

A node

Answer 1

A node is a machine – physical or virtual – on which kubernetes is installed. A node is a worker machine and this is were containers will be launched by kubernetes.

Question 2

A minion

Answer 2

A node

Question 3

A cluster

Answer 3

A cluster is a set of nodes grouped together. This way even if one node fails you have your application still accessible from the other nodes. Moreover having multiple nodes helps in sharing load as well.

Question 4

The master

Answer 4

The master is another node with Kubernetes installed in it, and is configured as a Master. The master watches over the nodes in the cluster and is responsible for the actual orchestration of containers on the worker nodes.

Question 5

Kubernetes components

Answer 5

An API Server. An ETCD service. A kubelet service. A Container Runtime, Controllers and Schedulers.

Question 6

The API server

Answer 6

The API server acts as the front-end for kubernetes. The users, management devices, Command line interfaces all talk to the API server to interact with the kubernetes cluster.

Question 7

ETCD

Answer 7

ETCD is a distributed reliable key-value store used by kubernetes to store all data used to manage the cluster. Think of it this way, when you have multiple nodes and multiple masters in your cluster, etcd stores all that information on all the nodes in the cluster in a distributed manner. ETCD is responsible for implementing locks within the cluster to ensure there are no conflicts between the Masters.

Question 8

The scheduler

Answer 8

The scheduler is responsible for distributing work or containers across multiple nodes. It looks for newly created containers and assigns them to Nodes.

Question 9

The controllers

Answer 9

The controllers are the brain behind orchestration. They are responsible for noticing and responding when nodes, containers or endpoints goes down. The controllers makes decisions to bring up new containers in such cases.

Question 10

The container runtime

Answer 10

The container runtime is the underlying software that is used to run containers. In our case it happens to be Docker.

Question 11

The kublet

Answer 11

And finally kubelet is the agent that runs on each node in the cluster. The agent is responsible for making sure that the containers are running on the nodes as expected. Kubelet is responsible for interacting with the master to provide health information of the worker node and carry out actions requested by the master on the worker nodes.

Question 12

The kube command line tool

Answer 12

ONE of the command line utilities used to deploy and manage applications on a kubernetes cluster, to get cluster information, get the status of nodes in the cluster and many other things.

Question 13

kubectl

Answer 13

ONE of the command line utilities used to deploy and manage applications on a kubernetes cluster, to get cluster information, get the status of nodes in the cluster and many other things.

Question 14

kubectl run

Answer 14

Deploy an application on the cluster.

Question 15

kubectl cluster-info

Answer 15

View information about the cluster.

Question 16

kubectl get pod

Answer 16

List all the nodes part of the cluster.

Question 17

Master vs worker nodes - worker.

Answer 17

Worker is were the containers are hosted.

A container runtime (Docker, Rocket, CRIO etc.)

Kubelet agent.

Question 18

Minikube vs Kubeadmin vs play-with-k8s.com

Answer 18

A single instance of Kubernetes in an All-in-one setup vs tool used to configure kubernetes in a multi-node setup. Online.

Question 19

Minikube

Answer 19

Minikube bundles all components into a single image providing us a pre-configured single node kubernetes cluster so we can get started in a matter of minutes.

The whole bundle is packaged into an ISO image and is available online for download. Minikube provides an executable command line utility that will AUTOMATICALLY download the ISO.

You must have a hypervisor installed, kubectl installed and minikube executable installed on your system.

Question 20

Kubeadmin

Answer 20

The tool helps us setup a multi node cluster with master and workers on separate machines.

Question 21

Kubeadmin 6 steps

Answer 21

1. You must have multiple systems or virtual machines created for configuring a cluster.
2. Install a container runtime on the hosts.
3. Install kubeadmin tool on all the nodes.
4. Initialize the Master server.
5. POD network
6. Join the worker nodes to the master node

Question 22

A pod

Answer 22

The containers are encapsulated into a Kubernetes object known as PODs. A POD is a single instance of an application. A POD is the smallest object, that you can create in kubernetes. oneToOne or oneToMany

Question 23

kubectl run nginx –image nginx

Dostęp do nginx z node / z poza

Answer 23

Tak / Nie

Question 24

2 deployment strategies

Answer 24

RECREATE UPDATE First destroy all old instances and deploy newer versions - the app is down an inaccessible to users.

ROLLING UPDATE (default) destroy old and deploy new one by one - the app never goes down.

Question 25

How can you update your deployment?

Answer 25

## Footnote **1. Update a definition file and run** kubectl apply -f my-updated-file.yml A new rollout is triggered and a new version of deployment is created. 2. kubectl set image deployment/my-app-deployment nginx=nginx:new-version (does not update yml)

Question 26

Upgrades / rollbacks under the hood

Answer 26

1. First deployment - a replica set with replicas( pods) created. 2. Upgrade - new replica set created and deployment of the new replica set and at the same time taking down the old pods (based on the strategy). 3. in case of rollback - You will have 2 replica sets - old and new. You can rollback kubectl rollout undo deployment/myapp-deployment

Question 27

**Deployment:** create get update status rollback

Answer 27

**kubectl** ## Footnote **create** -f my-deployment.yml / **run** my-pod --image=x **get deployments** **apply** -f my-deployment.yml **set image** deployment/my-deployment --image=x:1 **rollout status** / **history / undo** deployment/my-deployment

Question 28

NEtworking

Answer 28

In docker world addres assigned to a container, in Kubernetes to a POD. Each pod in a node get internal IP address. Kube creates internal network for a node e.g. **10.244.0.0** and all pods inthe node are attached to this. eg **10.244.0.2 , 3, 4. The internal IP can change?.** For a cluster you need a 3rp party sollution, so that each pods can communicate with each other without IP conflicts. e.g. Calico Kubernetes expects us to configuret the network in a cluster so that all nodes /pods/ container can communicate with each other without NAT

Question 29

Services

Answer 29

They enable loose coupling betwen microservices in our application. e.g. we have a web app in a pod. From inside the node we can curl the pod. From outside: NodePort service. Service listeins on node's port and forwards to pod http://:

Question 30

Service's types

Answer 30

**NodePort** : world to pod **ClusterIP**: virtual IP inside the cluster to enable communication between different services such as set of frontend servers and a set of backend servers. **LoadBalacer**: for use with cloud provider - e.g. to distribute the load between servers in the frontend tier.

Question 31

NodePort Service

Answer 31

The service is like a virtual addres inside the node. Inside the cluster it has it's onw IP address, called the cluster IP of the service **NODE PORT**\<-\>cIOTS:**PORT**\<-\>**TARGET PORT** 3200-32767

Question 32

my-service.yaml - root

Answer 32

apiVersion: v1 kind: Service metadata: name: my--service spec:

Question 33

my-service.yaml spec

Answer 33

spec: type: NodePort ports: - targetPort: 80 port: 80 nodePort: 3000 selctor: app: my-pod type: web-tier

Question 34

targetPort, port, nodePort - which are mandatory?

Answer 34

**port,** targetPort will be the same**,** nodePort will be random 30000 - 32767

Question 35

A service uses selector to know the pod: 1. what's with multiple pods in a node? 2. what's with multiple pods accross nodes?

Answer 35

1. acts as a load balacer - random 2. Kubernetes creates a service that spans all nodes: From outside you use : To sume up - no matter if a single /multiple pods / multiple nodes, you create the service exactly the same way.

Question 36

What's the use of a cluster IP

Answer 36

Kube provides us with an interface to access a group of pods. Layers can scale without impacting other layers. Plus - pods' ip's change so we should not rely on that communication. And we don't need to make a desision which pod shoud communicate whit which, between layers (random algorithm).

Question 37

clusterIP.yaml (all)

Answer 37

apiVerison: v1 kind: Service metadata: name: back-tier spec: type: ClusterIP ports: - targetPort: port: selector: app: my-app type: backedn-tier

Question 38

LoadBalancer

Answer 38

Without native cloud support acts as a node port. To load balance you may use either native or vm with eg,. nginx