CloudIQ

Kubernetes
 CloudIQ

Writing
YAML Files
Deploying Kubernetes Objects
 CloudIQ
POD Deployment
Containers

Control Plane
POD
POD Name: metadata.name+<Random#>
Deployment Object POD IP: 178.89.55.34
Scheduler Name POD Label app: corpwebsite
metadata.name

Worker Node 2
API Server
etcd
Calico CNI plugin
Deployment
Object
Containers
Control
Manager
POD
POD Name: metadata.name+<Random#>
POD IP: 172.36.23.25
POD Label app: corpwebsite

template: Worker Node 1

kubectl metadata:
Deployment Spec labels:
 CloudIQ
PODs are Grouped by LABELs
Containers
Replicas = 2
Control Plane
POD
POD Name: metadata.name+<Random#>
Deployment Object POD IP: 178.89.55.34
Scheduler Name
metadata.name POD Label app: corpwebsite

Worker Node 2
API Server
etcd

Deployment
Object
Containers
Control
Manager
POD
POD Name: metadata.name+<Random#>
Replicas Set POD IP: 172.36.23.25
Desired = 2
Current =2 POD Label app: corpwebsite
POD Label:
app: corpwebsite Worker Node 1
 CloudIQ
POD Network Access

Containers Containers

POD POD
POD IP: 172.36.23.25
POD IP: 178.89.55.34

Worker Node 2 Worker Node 1

Node IP Node IP
Cluster Network

● PODs can talk to each other within Kubernetes Cluster Network

● They are not exposed to External Network or Internet

 CloudIQ
POD Service: NodePort

Containers Containers

POD IP: 178.89.55.34 POD IP: 172.36.23.25

POD Label app: corpwebsite POD Label app: corpwebsite

Node IP Node IP

Kubernetes
Traffic Route: Service Object
Internet > Node IP > POD IP NodePort

Azure Load Balancer

 CloudIQ
We have to write 2 YAML Files here

POD Deployment YAML File POD Service YAML File

● For Deploying PODs ● For Providing External Access to PODs

● PODs Identified by POD LABEL ● PODs Identified by POD LABEL

How to write?
● Write all in a SINGLE YAML File
● Write 2 separate YAML Files (Recommended)
 CloudIQ
YAML Skeleton
kind: Deployment When writing YAML files for Kubernetes, there are four
apiVersion: required fields that must be present.

metadata: APIVersion
name:
labels: Deployment Identification
Kind
label:
Metadata
spec:
replicas: Specifications
strategy: Pod Behaviour
selector:
matchLabels:
label:

template:
metadata:
labels: Pod Identification
label:

spec:
containers:
- name: Container Identification
image:
 CloudIQ
POD Deployment
apiVersion: apps/v1
kind: Deployment

metadata: ● Deployment Name

name: deployment-corpwebsite

spec:
selector: ● How many Replicas?
matchLabels:
app: pod-corpwebsite ● Group these Replicas using the
replicas: 2
particular POD LABEL
template:
metadata:
labels:
● POD LABELs
app: pod-corpwebsite
env: dev

spec:
containers:
- name: container-corpwebsite ● Container Specs
image: tanvisinghny/ssl-website
ports:
- containerPort: 80
- containerPort: 443
 CloudIQ
Service Deployment
apiVersion: v1
kind: Service

metadata: ● Service Name

name: service-corpwebsite

spec:
type: NodePort ● Service Type: NodePort, Load
selector: Balancer
app: pod-corpwebsite

● nodePort - a static port assigned on

ports:
each the node
- nodePort: 30163
port: 443 ● port - port exposed internally in the
targetPort: 443 cluster
externalIPs:
● targetPort - the container port to
- 165.232.184.39 (workernode1 IP)
send requests to
- 139.59.19.74 (workernode2 IP)

● Worker Node IPS where the PODs are

running
ports: CloudIQ
- nodePort: 30163
port: 443
targetPort: 443 https://www.krishnaazure.xyz:444
139.59.19.74

nodePort: 30000 - 32767

Containers Containers Containers

port: 443 port: 443 port: 443

nodePort: 30163 nodePort: 32000 nodePort: 32005

165.232.184.39 (443)

https://www.cloudiq.in targetPort: 443

https://www.krishnaazure.xyz targetPort: 443

 CloudIQ
POD Resource Limits
Kubernetes doesn’t provide default resource limits
out-of-the-box

Limits = Max amount of a resource (RAM or CPU)

Requests = Min guaranteed amount of a resource (RAM or

CPU)

resources: resources:
limits: limits:
memory: 200Mi memory: 2G
cpu: 300m cpu: 3
requests: requests: CPU represents computing processing time, measured in
memory: 100Mi memory: 1G
cpu: 100m cores.
cpu: 1
● You can use millicores (m) to represent smaller
amounts than a core (e.g., 500m would be half a
Memory is measured in Kubernetes in bytes:
core)
You can use, E, P, T, G, M, k to represent Exabyte, Petabyte,
Terabyte, Gigabyte, Megabyte and kilobyte ● The minimum amount is 1m

You can define Mebibytes using Mi, as well as the rest as Ei, Pi, ● A Node might have more than one core available, so
Ti (e.g., 500Mi) requesting CPU > 1 is possible
 CloudIQ
YAMl File with Resource Limits
apiVersion: apps/v1
kind: Deployment
metadata:
name: deployment-corpwebsite
spec:
selector:
matchLabels:
app: pod-corpwebsite
replicas: 2
template:
metadata:
labels:
app: pod-corpwebsite
env: dev
spec:
containers:
- name: container-corpwebsite
image: tanvisinghny/ssl-website
ports:
- containerPort: 80
- containerPort: 443
resources:
limits:
memory: 200Mi
cpu: 300m
requests:
memory: 100Mi
cpu: 100m
 CloudIQ
POD Resource Limits
Verifying Resource Limits set

kubectl describe deployment <deployment-name>

kubectl describe deployment <pod-name>

Verifying Resource Limits inside the pod

Go inside POD:

kubectl exec -it <pod-name> -n <namespace> -- /bin/bash

Get CPU, Memory Stats

cd /sys/fs/cgroup/

Find various statistics based on files described in that

folder
 CloudIQ
Built-in Kubernetes Metrics
Install Metrics-Server Components

curl -LO https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

Edit the file and ADD the metrics:

sudo nano components.yaml

Once done, apply the changes

kubectl apply -f components.yaml

 CloudIQ
Verify if Metrics-Server is Running
linuxadmin@master:~$ kubectl get pods -n kube-system

NAME READY STATUS RESTARTS AGE

calico-kube-controllers-6c99c8747f-tgc6w 1/1 Running 0 36h

calico-node-dgjww 1/1 Running 0 36h

calico-node-hzzbg 1/1 Running 0 36h

—--

metrics-server-f94d65847-vwq2s 1/1 Running 0 27m

 CloudIQ
Test Metrics Server Installation
Display the resource utilization for each node in your cluster, including CPU and memory usage

linuxadmin@master:~$ kubectl top nodes

NAME CPU(cores) CPU% MEMORY(bytes) MEMORY%

master 138m 6% 2235Mi 58%

node1 49m 2% 1636Mi 42%

node2 40m 2% 1601Mi 41%

 CloudIQ
Test Metrics Server Installation
View pods resource utilization of your current namespace

linuxadmin@master:~$ kubectl top pod

NAME CPU(cores) MEMORY(bytes)

deployment-corpwebsite-7675c48cc6-6zn5d 1m 7Mi

deployment-corpwebsite-7675c48cc6-rl4sh 1m 7Mi
 CloudIQ
Test Metrics Server Installation
view pods resource utilization of your current namespace

linuxadmin@master:~$ kubectl top pod

NAME CPU(cores) MEMORY(bytes)

deployment-corpwebsite-7675c48cc6-6zn5d 1m 7Mi

deployment-corpwebsite-7675c48cc6-rl4sh 1m 7Mi