cka-prep/notes at master · malikons/cka-prep · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
etcd
=====

kubectl exec etcd-master -n kube-system etcdctl get / --prefix -keys-only

*********** MASTER NODE ************

kube-apiserver (on Master)
==============
kubectl --> talks to kube-apiserver
can also use API directly --> curl -X POST /api/v1/namespaces/default/pods ...[other]

can be downloaded as a binary from k8s release
kube-apiserver.service

kubectl get pods -n kube-system
kube-apiserver-master (on master node)
cat /etc/kubernets/manifests/kube-apiserver.yaml
cat /etc/systemd/system/kube-apiserver.service
ps -aux | grep kube-apiserver

kube-controller-manager
=======================
1. Watch Status
2. Remediate Situations

-- Node-Controller----->(5sec)---->kube-apiserver------->[nodes]
(wait for 40 seconds once node goes down and marks are unreachable - then waits for 5 mins to come back up - if not, then it starts evacuating the pods and run them on available nodes)
POD Eviction Timeout = 5m

-- replication-controllers
monitor is desired number of pods are running at all times

many more controllers available

kube-controller-manager.service

kubectl get pods -n kube-system

cat /etc/kubernetes/manifests/kube-controller-manager.yaml
cat /etc/systemd/system/kube-controller-manager.service

ps -aux | grep kube-controller-manager

kube-scheduler
===============
decides what pods go to which node (doesn't do it himself, that's kubelet job)
1. Filter Nodes (eg: based on resource requirements, affinity rules, etc.)
2. Rank Nodes (eg: how many resources will be left after placing a pod - higher wins)

cat /etc/kubernetes/manifests/kube-scheduler.yaml
cat /etc/systemd/
ps -aux

*********** WORKER NODE ************

Kubelet:
=========

kubeadm does not deploy kubelet

ps -aux | grep kubelet

kube-proxy
===========
every pod can reach every pod - using virtual network - all pods connect to this network
many solution available for deploying this
uses iptables rules on every node to tell where the traffic needs to go for particular services

can download from k8s release page

kubectl get pods -n kube-system
deployed as deamonset (every node will have at least 1 pod)
kubectl get deamonset -n kube-system

PODs
=======
containers are encapsulated inside pods in k8s
do not add containers in the same pod to scale application


YAML
=====
(indention matters - should be consistent)

mandatory fields:
------------------
apiVersion: --> version of k8s api of what we are creating
kind: --> Pod (e.g.: ReplicaSet, Deployment, etc.)
metadata:
  name: myapp-pod
  labels:
     app: myapp
     type: front-end --> (to make it easier to find pods based on their function)
spec:
  containers:  --> List/Array
    - name: nginx-container ("-" to tell this is first item in the list)
      image: nginx

kubectl describe pod myapp-pod

PODS Demo
=========
- writing pod.yaml file

Generate pod yaml:
==================
kubectl run nginx --image=nginx --dry-run -o yaml

Generate deployment YAML:
=========================
kubectl create deployment --image=httpd:2.4-alpine httpd-frontend --dry-run -o yaml > deployment-httpd-frontend.yaml


master $ history
    1  clear
    2  RED='\033[0;31m'
    3  NC='\033[0m'
    4  trap 'rm -rf /tmp/load-quiz.sh ; rm -rf /tmp/wait-script.sh' SIGINT SIGQUIT ERR SIGTSTP
    5  clear
    6  while [ ! -f /tmp/wait-script.sh ] ; do clear; echo "Waiting for environment to load";   sleep .2; done
    7  chmod 755 /tmp/wait-script.sh; /tmp/wait-script.sh
    8  clear
    9  kubectl get pods
   10  kubectl get rs
   11  kubectl get deployments
   12  kubectl get rs
   13  kubectl get pods
   14  kubectl describe pod frontend-deployment-55c859888c-hxk4b
   15  kubectl create -f deployment-definition-1.yaml
   16  vim deployment-definition-1.yaml
   17  kubectl create -f deployment-definition-1.yaml
   18  kubectl get deployments
   19  kubectl create deployment --image=httpd:2.4-alpine httpd-frontend --dry-run -o yaml > deployment-httpd-frontend.yaml
   20  cat deployment-httpd-frontend.yaml
   21  vim deployment-httpd-frontend.yaml
   22  kubectl create deployment -f deployment-httpd-frontend.yaml
   23  kubectl create -f deployment-httpd-frontend.yaml
   24  kubectl get deployment
   25  kubectl run nginx --image=nginx --dry-run -o yaml
   26  history