Unix Server Solutions: kubernetes Pod Scheduling

===================

Deployment

=================

1.) Deployment without any nodeName or nodeSelector, pod will spread among all of the available worker node as by default

schedular use spread policy for pod placment.

anuj@maste01:~$ cat nginx-deployment.yaml

apiVersion: apps/v1

kind: Deployment

metadata:

labels:

app: nginx

spec:

strategy:

type: RollingUpdate

rollingUpdate:

maxSurge: 2

maxUnavailable: 0

revisionHistoryLimit: 4

replicas: 8

minReadySeconds: 20

selector:

matchLabels:

app: nginx

template:

metadata:

labels:

app: nginx

spec:

containers:

- name: nginx

image: nginx:1.14.2

ports:

- containerPort: 80

anuj@maste01:~$

anuj@maste01:~$ kubectl apply -f nginx-deployment.yaml

deployment.apps/nginx-deployment created

//* Now you can see nginx pod placed on both worker node *//

anuj@maste01:~$

anuj@maste01:~$ kubectl get po -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-66b6c48dd5-6qbg5 1/1 Running 0 6s 10.32.0.7 worker01 <none> <none>

nginx-deployment-66b6c48dd5-6wl7j 1/1 Running 0 6s 10.40.0.9 worker02 <none> <none>

nginx-deployment-66b6c48dd5-99xnc 1/1 Running 0 6s 10.40.0.8 worker02 <none> <none>

nginx-deployment-66b6c48dd5-hprz5 1/1 Running 0 6s 10.40.0.7 worker02 <none> <none>

nginx-deployment-66b6c48dd5-j9s72 1/1 Running 0 6s 10.32.0.9 worker01 <none> <none>

nginx-deployment-66b6c48dd5-r7lmf 1/1 Running 0 6s 10.32.0.8 worker01 <none> <none>

nginx-deployment-66b6c48dd5-rcqd5 1/1 Running 0 6s 10.40.0.6 worker02 <none> <none>

nginx-deployment-66b6c48dd5-xsjpq 1/1 Running 0 6s 10.40.0.5 worker02 <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 44h 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 44h 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 44h 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 44h 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 16h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 16h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running 0 16h 10.32.0.4 worker01 <none> <none>

============== pod placement using nodeName ============

using nodeName keyword, just before the target node where you want that your pod should run on that particular node only.

you can see in following output nginx pod are running on both worker nodes.

Now lets do some update to above used deployment config file so that all nginx-deployment pod will run only worker02 only

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-66b6c48dd5-6qbg5 1/1 Running 0 5m27s 10.32.0.7 worker01 <none> <none> <===== worker01

nginx-deployment-66b6c48dd5-6wl7j 1/1 Running 0 5m27s 10.40.0.9 worker02 <none> <none> <===== worker02

nginx-deployment-66b6c48dd5-99xnc 1/1 Running 0 5m27s 10.40.0.8 worker02 <none> <none> <===== worker02

nginx-deployment-66b6c48dd5-hprz5 1/1 Running 0 5m27s 10.40.0.7 worker02 <none> <none> <===== worker02

nginx-deployment-66b6c48dd5-j9s72 1/1 Running 0 5m27s 10.32.0.9 worker01 <none> <none> <===== worker01

nginx-deployment-66b6c48dd5-r7lmf 1/1 Running 0 5m27s 10.32.0.8 worker01 <none> <none> <===== worker01

nginx-deployment-66b6c48dd5-rcqd5 1/1 Running 0 5m27s 10.40.0.6 worker02 <none> <none> <===== worker02

nginx-deployment-66b6c48dd5-xsjpq 1/1 Running 0 5m27s 10.40.0.5 worker02 <none> <none> <===== worker02

vote-5d548444c5-5tgbf 1/1 Running 0 44h 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 44h 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 44h 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 44h 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 16h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 16h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running 0 16h 10.32.0.4 worker01 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ cat nginx-deployment.yaml

apiVersion: apps/v1

kind: Deployment

metadata:

labels:

app: nginx

spec:

strategy:

type: RollingUpdate

rollingUpdate:

maxSurge: 2

maxUnavailable: 0

revisionHistoryLimit: 4

replicas: 8

minReadySeconds: 20

selector:

matchLabels:

app: nginx

template:

metadata:

labels:

app: nginx

spec:

containers:

- name: nginx

image: nginx:1.14.2

ports:

- containerPort: 80

nodeName: worker02 <================ Secheduler is by pass and pod will be scheduled on node which is specified

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-66b6c48dd5-6qbg5 1/1 Running 0 8m25s 10.32.0.7 worker01 <none> <none>

nginx-deployment-66b6c48dd5-6wl7j 1/1 Running 0 8m25s 10.40.0.9 worker02 <none> <none>

nginx-deployment-66b6c48dd5-99xnc 1/1 Running 0 8m25s 10.40.0.8 worker02 <none> <none>

nginx-deployment-66b6c48dd5-hprz5 1/1 Running 0 8m25s 10.40.0.7 worker02 <none> <none>

nginx-deployment-66b6c48dd5-j9s72 1/1 Running 0 8m25s 10.32.0.9 worker01 <none> <none>

nginx-deployment-66b6c48dd5-r7lmf 1/1 Running 0 8m25s 10.32.0.8 worker01 <none> <none>

nginx-deployment-66b6c48dd5-rcqd5 1/1 Running 0 8m25s 10.40.0.6 worker02 <none> <none>

nginx-deployment-66b6c48dd5-xsjpq 1/1 Running 0 8m25s 10.40.0.5 worker02 <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 44h 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 44h 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 44h 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 44h 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 16h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 16h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running 0 16h 10.32.0.4 worker01 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl apply -f nginx-deployment.yaml

deployment.apps/nginx-deployment configured

anuj@maste01:~$

//* All pods deleted will be redploy to worker02 *//

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-66b6c48dd5-6qbg5 1/1 Running 0 9m16s 10.32.0.7 worker01 <none> <none>

nginx-deployment-66b6c48dd5-6wl7j 1/1 Terminating 0 9m16s 10.40.0.9 worker02 <none> <none>

nginx-deployment-66b6c48dd5-hprz5 1/1 Terminating 0 9m16s 10.40.0.7 worker02 <none> <none>

nginx-deployment-66b6c48dd5-j9s72 1/1 Running 0 9m16s 10.32.0.9 worker01 <none> <none>

nginx-deployment-66b6c48dd5-r7lmf 1/1 Running 0 9m16s 10.32.0.8 worker01 <none> <none>

nginx-deployment-66b6c48dd5-xsjpq 1/1 Running 0 9m16s 10.40.0.5 worker02 <none> <none>

nginx-deployment-845f7554fc-57xkk 0/1 ContainerCreating 0 1s <none> worker02 <none> <none>

nginx-deployment-845f7554fc-fztw8 1/1 Running 0 24s 10.40.0.8 worker02 <none> <none>

nginx-deployment-845f7554fc-gpgms 0/1 ContainerCreating 0 1s <none> worker02 <none> <none>

nginx-deployment-845f7554fc-mp2rj 1/1 Running 0 24s 10.40.0.6 worker02 <none> <none>

nginx-deployment-845f7554fc-np8d4 1/1 Running 0 46s 10.40.0.11 worker02 <none> <none>

nginx-deployment-845f7554fc-zt7lk 1/1 Running 0 46s 10.40.0.10 worker02 <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 44h 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 44h 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 44h 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 44h 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 16h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 16h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running 0 16h 10.32.0.4 worker01 <none> <none>

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-66b6c48dd5-6qbg5 1/1 Running 0 9m19s 10.32.0.7 worker01 <none> <none>

nginx-deployment-66b6c48dd5-6wl7j 0/1 Terminating 0 9m19s 10.40.0.9 worker02 <none> <none>

nginx-deployment-66b6c48dd5-hprz5 0/1 Terminating 0 9m19s 10.40.0.7 worker02 <none> <none>

nginx-deployment-66b6c48dd5-j9s72 1/1 Running 0 9m19s 10.32.0.9 worker01 <none> <none>

nginx-deployment-66b6c48dd5-r7lmf 1/1 Running 0 9m19s 10.32.0.8 worker01 <none> <none>

nginx-deployment-66b6c48dd5-xsjpq 1/1 Running 0 9m19s 10.40.0.5 worker02 <none> <none>

nginx-deployment-845f7554fc-57xkk 1/1 Running 0 4s 10.40.0.9 worker02 <none> <none>

nginx-deployment-845f7554fc-fztw8 1/1 Running 0 27s 10.40.0.8 worker02 <none> <none>

nginx-deployment-845f7554fc-gpgms 1/1 Running 0 4s 10.40.0.7 worker02 <none> <none>

nginx-deployment-845f7554fc-mp2rj 1/1 Running 0 27s 10.40.0.6 worker02 <none> <none>

nginx-deployment-845f7554fc-np8d4 1/1 Running 0 49s 10.40.0.11 worker02 <none> <none>

nginx-deployment-845f7554fc-zt7lk 1/1 Running 0 49s 10.40.0.10 worker02 <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 44h 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 44h 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 44h 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 44h 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 16h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 16h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running

//* Can you see all nginx-deployment pods are now running on worker02 and no pod is present on worker01

beacuse under deployment config you have mentioned dedicated node to host all pod of nginx-deployment,

hence schedular will not come in to the picture to check where to place the pod *//

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-845f7554fc-57xkk 1/1 Running 0 2m57s 10.40.0.9 worker02 <none> <none>

nginx-deployment-845f7554fc-bj6mf 1/1 Running 0 2m17s 10.40.0.5 worker02 <none> <none>

nginx-deployment-845f7554fc-fztw8 1/1 Running 0 3m20s 10.40.0.8 worker02 <none> <none>

nginx-deployment-845f7554fc-gpgms 1/1 Running 0 2m57s 10.40.0.7 worker02 <none> <none>

nginx-deployment-845f7554fc-mp2rj 1/1 Running 0 3m20s 10.40.0.6 worker02 <none> <none>

nginx-deployment-845f7554fc-np8d4 1/1 Running 0 3m42s 10.40.0.11 worker02 <none> <none>

nginx-deployment-845f7554fc-wffjg 1/1 Running 0 2m17s 10.40.0.12 worker02 <none> <none>

nginx-deployment-845f7554fc-zt7lk 1/1 Running 0 3m42s 10.40.0.10 worker02 <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 44h 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 44h 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 44h 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 44h 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 16h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 16h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running 0 16h 10.32.0.4 worker01 <none> <none>

anuj@maste01:~$

===========================

nodeSelector/node affinity

===========================

//* Now let say you have 6 worker node and you have grouped them in some types like UAT DEV PROD and each type have 2 worker node.

now you want to run you application on only nodes those in UAT group(type) note on other nodes.

For that we have to need following.

1.) label the nodes

2.) under the deployment config have to use nodeSelector:

env: UAT

anuj@maste01:~$ cat nginx-deployment.yaml

apiVersion: apps/v1

kind: Deployment

metadata:

labels:

app: nginx

spec:

strategy:

type: RollingUpdate

rollingUpdate:

maxSurge: 2

maxUnavailable: 0

revisionHistoryLimit: 4

replicas: 8

minReadySeconds: 20

selector:

matchLabels:

app: nginx

template:

metadata:

labels:

app: nginx

spec:

containers:

- name: nginx

image: nginx:1.14.2

ports:

- containerPort: 80

nodeSelector:

env: uat

anuj@maste01:~$

anuj@maste01:~$ kubectl apply -f nginx-deployment.yaml

deployment.apps/nginx-deployment configured

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-59bdc8fccc-bjs2g 1/1 Running 0 45s 10.32.0.9 worker01 <none> <none>

nginx-deployment-59bdc8fccc-bn4r9 1/1 Running 0 66s 10.32.0.7 worker01 <none> <none>

nginx-deployment-59bdc8fccc-cvm2h 0/1 ContainerCreating 0 0s <none> worker01 <none> <none>

nginx-deployment-59bdc8fccc-df4fl 1/1 Running 0 66s 10.32.0.8 worker01 <none> <none>

nginx-deployment-59bdc8fccc-gtwgd 1/1 Running 0 22s 10.32.0.12 worker01 <none> <none>

nginx-deployment-59bdc8fccc-nc8xd 1/1 Running 0 22s 10.32.0.11 worker01 <none> <none>

nginx-deployment-59bdc8fccc-nrcbj 0/1 ContainerCreating 0 0s <none> worker01 <none> <none>

nginx-deployment-59bdc8fccc-rqxbj 1/1 Running 0 45s 10.32.0.10 worker01 <none> <none>

nginx-deployment-845f7554fc-5fqtl 1/1 Running 0 2m54s 10.40.0.6 worker02 <none> <none>

nginx-deployment-845f7554fc-h4kfz 1/1 Terminating 0 2m32s 10.40.0.8 worker02 <none> <none>

nginx-deployment-845f7554fc-mklzx 1/1 Terminating 0 2m32s 10.40.0.7 worker02 <none> <none>

nginx-deployment-845f7554fc-nzkvp 1/1 Running 0 2m54s 10.40.0.5 worker02 <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 2d 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 2d 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 2d 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 2d 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 20h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 20h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running 0 20h 10.32.0.4 worker01 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-59bdc8fccc-bjs2g 1/1 Running 0 2m30s 10.32.0.9 worker01 <none> <none>

nginx-deployment-59bdc8fccc-bn4r9 1/1 Running 0 2m51s 10.32.0.7 worker01 <none> <none>

nginx-deployment-59bdc8fccc-cvm2h 1/1 Running 0 105s 10.32.0.13 worker01 <none> <none>

nginx-deployment-59bdc8fccc-df4fl 1/1 Running 0 2m51s 10.32.0.8 worker01 <none> <none>

nginx-deployment-59bdc8fccc-gtwgd 1/1 Running 0 2m7s 10.32.0.12 worker01 <none> <none>

nginx-deployment-59bdc8fccc-nc8xd 1/1 Running 0 2m7s 10.32.0.11 worker01 <none> <none>

nginx-deployment-59bdc8fccc-nrcbj 1/1 Running 0 105s 10.32.0.14 worker01 <none> <none>

nginx-deployment-59bdc8fccc-rqxbj 1/1 Running 0 2m30s 10.32.0.10 worker01 <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 2d 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 2d 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 2d 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 2d 10.40.0.2 worker02 <none> <none>

web-699689dbd-8lhkr 1/1 Running 0 20h 10.32.0.2 worker01 <none> <none>

web-699689dbd-j6s87 1/1 Running 0 20h 10.32.0.3 worker01 <none> <none>

web-699689dbd-mt22j 1/1 Running 0 20h 10.32.0.4 worker01 <none> <none>

anuj@maste01:~$

//* Just see all nginx pods are running on worker01 that has label set env=uat *//

anuj@maste01:~$ kubectl describe pod/nginx-deployment-59bdc8fccc-cvm2h

Name: nginx-deployment-59bdc8fccc-cvm2h

Namespace: default

Priority: 0

Node: worker01/192.168.93.129

Start Time: Fri, 29 Apr 2022 02:35:42 -0700

Labels: app=nginx

pod-template-hash=59bdc8fccc

Annotations: <none>

Status: Running

IP: 10.32.0.13

IPs:

IP: 10.32.0.13

Controlled By: ReplicaSet/nginx-deployment-59bdc8fccc

Containers:

nginx:

Container ID: docker://ee430bfbc07a199214bbc0692c89d498cf996441d45e2bb9ea0c7a725ec74fc3

Image: nginx:1.14.2

Image ID: docker-pullable://nginx@sha256:f7988fb6c02e0ce69257d9bd9cf37ae20a60f1df7563c3a2a6abe24160306b8d

Port: 80/TCP

Host Port: 0/TCP

State: Running

Started: Fri, 29 Apr 2022 02:35:44 -0700

Ready: True

Restart Count: 0

Environment: <none>

Mounts:

/var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-7tzjk (ro)

Conditions:

Type Status

Initialized True

Ready True

ContainersReady True

PodScheduled True

Volumes:

kube-api-access-7tzjk:

Type: Projected (a volume that contains injected data from multiple sources)

TokenExpirationSeconds: 3607

ConfigMapName: kube-root-ca.crt

ConfigMapOptional: <nil>

DownwardAPI: true

QoS Class: BestEffort

Node-Selectors: env=uat <=========================== nodeSelector is set for pod env=uat so it will schedule the pod on node

that has env=uat label is set.

Tolerations: node.kubernetes.io/not-ready:NoExecute op=Exists for 300s

node.kubernetes.io/unreachable:NoExecute op=Exists for 300s

Events:

Type Reason Age From Message

---- ------ ---- ---- -------

Normal Scheduled 8m55s default-scheduler Successfully assigned default/nginx-deployment-59bdc8fccc-cvm2h to worker01 <=== Scheduler

scheduled pod on worker01

Normal Pulled 8m55s kubelet Container image "nginx:1.14.2" already present on machine

Normal Created 8m55s kubelet Created container nginx

Normal Started 8m54s kubelet Started container nginx

anuj@maste01:~$

//* refer Follolwing output and you can see how replicaset create pod with app=nginx label using pod template *//

anuj@maste01:~$ kubectl get rs

NAME DESIRED CURRENT READY AGE

nginx-deployment-59bdc8fccc 8 8 8 17m

nginx-deployment-66b6c48dd5 0 0 0 4h5m

nginx-deployment-845f7554fc 0 0 0 3h56m

vote-5d548444c5 4 4 4 2d

vote-7fdc744cbf 0 0 0 2d

web-55d678bd48 0 0 0 20h

web-699689dbd 3 3 3 20h

anuj@maste01:~$

anuj@maste01:~$ kubectl describe rs/nginx-deployment-59bdc8fccc

Name: nginx-deployment-59bdc8fccc

Namespace: default

Selector: app=nginx,pod-template-hash=59bdc8fccc

Labels: app=nginx

pod-template-hash=59bdc8fccc

Annotations: deployment.kubernetes.io/desired-replicas: 8

deployment.kubernetes.io/max-replicas: 10

deployment.kubernetes.io/revision: 5

deployment.kubernetes.io/revision-history: 3

Controlled By: Deployment/nginx-deployment

Replicas: 8 current / 8 desired

Pods Status: 8 Running / 0 Waiting / 0 Succeeded / 0 Failed

Pod Template:

Labels: app=nginx

pod-template-hash=59bdc8fccc

Containers:

nginx:

Image: nginx:1.14.2

Port: 80/TCP

Host Port: 0/TCP

Environment: <none>

Mounts: <none>

Volumes: <none>

Events:

Type Reason Age From Message

---- ------ ---- ---- -------

Normal SuccessfulCreate 17m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-x499j

Normal SuccessfulCreate 17m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-ln9wf

Normal SuccessfulCreate 16m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-zgpzw

Normal SuccessfulCreate 16m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-97kpt

Normal SuccessfulCreate 16m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-kjpjb

Normal SuccessfulCreate 16m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-j62qz

Normal SuccessfulCreate 16m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-5zqcg

Normal SuccessfulCreate 16m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-mm4nl

Normal SuccessfulDelete 14m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-5zqcg

Normal SuccessfulDelete 14m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-mm4nl

Normal SuccessfulDelete 13m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-kjpjb

Normal SuccessfulDelete 13m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-j62qz

Normal SuccessfulDelete 13m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-97kpt

Normal SuccessfulDelete 13m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-zgpzw

Normal SuccessfulDelete 13m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-x499j

Normal SuccessfulDelete 13m replicaset-controller Deleted pod: nginx-deployment-59bdc8fccc-ln9wf

Normal SuccessfulCreate 12m replicaset-controller Created pod: nginx-deployment-59bdc8fccc-bn4r9

Normal SuccessfulCreate 11m (x7 over 12m) replicaset-controller (combined from similar events): Created pod: nginx-deployment-59bdc8fccc-cvm2h

anuj@maste01:~$ kubectl get pod -l app=nginx

NAME READY STATUS RESTARTS AGE

nginx-deployment-59bdc8fccc-bjs2g 1/1 Running 0 12m

nginx-deployment-59bdc8fccc-bn4r9 1/1 Running 0 13m

nginx-deployment-59bdc8fccc-cvm2h 1/1 Running 0 12m

nginx-deployment-59bdc8fccc-df4fl 1/1 Running 0 13m

nginx-deployment-59bdc8fccc-gtwgd 1/1 Running 0 12m

nginx-deployment-59bdc8fccc-nc8xd 1/1 Running 0 12m

nginx-deployment-59bdc8fccc-nrcbj 1/1 Running 0 12m

nginx-deployment-59bdc8fccc-rqxbj 1/1 Running 0 12m

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod --show-labels

NAME READY STATUS RESTARTS AGE LABELS

nginx-deployment-59bdc8fccc-bjs2g 1/1 Running 0 13m app=nginx,pod-template-hash=59bdc8fccc

nginx-deployment-59bdc8fccc-bn4r9 1/1 Running 0 13m app=nginx,pod-template-hash=59bdc8fccc

nginx-deployment-59bdc8fccc-cvm2h 1/1 Running 0 12m app=nginx,pod-template-hash=59bdc8fccc

nginx-deployment-59bdc8fccc-df4fl 1/1 Running 0 13m app=nginx,pod-template-hash=59bdc8fccc

nginx-deployment-59bdc8fccc-gtwgd 1/1 Running 0 12m app=nginx,pod-template-hash=59bdc8fccc

nginx-deployment-59bdc8fccc-nc8xd 1/1 Running 0 12m app=nginx,pod-template-hash=59bdc8fccc

nginx-deployment-59bdc8fccc-nrcbj 1/1 Running 0 12m app=nginx,pod-template-hash=59bdc8fccc

nginx-deployment-59bdc8fccc-rqxbj 1/1 Running 0 13m app=nginx,pod-template-hash=59bdc8fccc

vote-5d548444c5-5tgbf 1/1 Running 0 2d app=python,pod-template-hash=5d548444c5,role=vote,version=v3

vote-5d548444c5-rb6rz 1/1 Running 0 2d app=python,pod-template-hash=5d548444c5,role=vote,version=v3

vote-5d548444c5-rsg4x 1/1 Running 0 2d app=python,pod-template-hash=5d548444c5,role=vote,version=v3

vote-5d548444c5-s5kps 1/1 Running 0 2d app=python,pod-template-hash=5d548444c5,role=vote,version=v3

web-699689dbd-8lhkr 1/1 Running 0 20h app=web,pod-template-hash=699689dbd

web-699689dbd-j6s87 1/1 Running 0 20h app=web,pod-template-hash=699689dbd

web-699689dbd-mt22j 1/1 Running 0 20h app=web,pod-template-hash=699689dbd

anuj@maste01:~$

//* now try to drain worker01 and see all pods those were meant to run on env=uat should go on pending state *//

as they will not get any node that has env=uat label is set

anuj@maste01:~$ kubectl drain worker01 --ignore-daemonsets

node/worker01 already cordoned

WARNING: ignoring DaemonSet-managed Pods: kube-system/kube-proxy-7dlzd, kube-system/weave-net-qmkzn

evicting pod kube-system/coredns-558bd4d5db-zz7lx

evicting pod default/nginx-deployment-59bdc8fccc-nc8xd

evicting pod default/nginx-deployment-59bdc8fccc-bjs2g

evicting pod default/nginx-deployment-59bdc8fccc-bn4r9

evicting pod default/nginx-deployment-59bdc8fccc-cvm2h

evicting pod default/nginx-deployment-59bdc8fccc-df4fl

evicting pod default/nginx-deployment-59bdc8fccc-gtwgd

evicting pod default/web-699689dbd-8lhkr

evicting pod default/nginx-deployment-59bdc8fccc-nrcbj

evicting pod default/nginx-deployment-59bdc8fccc-rqxbj

evicting pod default/web-699689dbd-mt22j

evicting pod default/web-699689dbd-j6s87

evicting pod kube-system/coredns-558bd4d5db-jhxck

I0429 03:29:22.578066 1093416 request.go:668] Waited for 1.0612795s due to client-side throttling, not priority and fairness, request: GET:https://192.168.93.128:6443/api/v1/namespaces/kube-system/pods/coredns-558bd4d5db-jhxck

pod/nginx-deployment-59bdc8fccc-cvm2h evicted

pod/web-699689dbd-j6s87 evicted

pod/nginx-deployment-59bdc8fccc-bjs2g evicted

pod/nginx-deployment-59bdc8fccc-nc8xd evicted

pod/web-699689dbd-mt22j evicted

pod/web-699689dbd-8lhkr evicted

pod/nginx-deployment-59bdc8fccc-gtwgd evicted

pod/nginx-deployment-59bdc8fccc-bn4r9 evicted

pod/nginx-deployment-59bdc8fccc-nrcbj evicted

pod/nginx-deployment-59bdc8fccc-rqxbj evicted

pod/nginx-deployment-59bdc8fccc-df4fl evicted

pod/coredns-558bd4d5db-zz7lx evicted

pod/coredns-558bd4d5db-jhxck evicted

node/worker01 evicted

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-59bdc8fccc-4pncp 0/1 Pending 0 14s <none> <none> <none> <none>

nginx-deployment-59bdc8fccc-6qb6p 0/1 Pending 0 14s <none> <none> <none> <none>

nginx-deployment-59bdc8fccc-8nscb 0/1 Pending 0 13s <none> <none> <none> <none>

nginx-deployment-59bdc8fccc-9cgwt 0/1 Pending 0 14s <none> <none> <none> <none>

nginx-deployment-59bdc8fccc-f7mfz 0/1 Pending 0 14s <none> <none> <none> <none>

nginx-deployment-59bdc8fccc-l6q7q 0/1 Pending 0 14s <none> <none> <none> <none>

nginx-deployment-59bdc8fccc-ljmfn 0/1 Pending 0 14s <none> <none> <none> <none>

nginx-deployment-59bdc8fccc-rpj9v 0/1 Pending 0 14s <none> <none> <none> <none>

vote-5d548444c5-5tgbf 1/1 Running 0 2d1h 10.40.0.4 worker02 <none> <none>

vote-5d548444c5-rb6rz 1/1 Running 0 2d1h 10.40.0.1 worker02 <none> <none>

vote-5d548444c5-rsg4x 1/1 Running 0 2d1h 10.40.0.3 worker02 <none> <none>

vote-5d548444c5-s5kps 1/1 Running 0 2d1h 10.40.0.2 worker02 <none> <none>

web-699689dbd-85kqx 0/1 Pending 0 14s <none> <none> <none> <none>

web-699689dbd-9p6w5 0/1 Pending 0 14s <none> <none> <none> <none>

web-699689dbd-k8nnz 0/1 Pending 0 14s <none> <none> <none> <none>

anuj@maste01:~$

// * Now lets see the logs of any pending pod and you will find node(s) didn't match Pod's node affinity/selector *//

anuj@maste01:~$ kubectl describe pod/nginx-deployment-59bdc8fccc-l6q7q | tail

DownwardAPI: true

QoS Class: BestEffort

Node-Selectors: env=uat

Tolerations: node.kubernetes.io/not-ready:NoExecute op=Exists for 300s

node.kubernetes.io/unreachable:NoExecute op=Exists for 300s

Events:

Type Reason Age From Message

---- ------ ---- ---- -------

Warning FailedScheduling 2m22s default-scheduler 0/3 nodes are available: 1 node(s) didn't match Pod's node affinity/selector, 1 node(s) had taint {node-role.kubernetes.io/master: }, that the pod didn't tolerate, 1 node(s) were unschedulable.

Warning FailedScheduling 2m21s default-scheduler 0/3 nodes are available: 1 node(s) didn't match Pod's node affinity/selector, 1 node(s) had taint {node-role.kubernetes.io/master: }, that the pod didn't tolerate, 1 node(s) were unschedulable.

anuj@maste01:~$

//* Now lets see the worker01 node status, its marked as unschedulable, hence no pod can be run on it *//

lets uncordon worker01 and all pods those are meant to run on this should come up by its own as they will get the node that have

env=uat set

anuj@maste01:~$ kubectl get node

NAME STATUS ROLES AGE VERSION

maste01 Ready control-plane,master 24d v1.21.0

worker01 Ready,SchedulingDisabled <none> 23d v1.21.0

worker02 Ready <none> 23d v1.21.0

anuj@maste01:~$ kubectl uncordon worker01

node/worker01 uncordoned

anuj@maste01:~$

//** see container start coming up by its own on worker01, hence its prove that node that have nodeSelector env=uat is set in

its deployment on that node pod will run not on any other nodes **//

anuj@maste01:~$ kubectl get pod

NAME READY STATUS RESTARTS AGE

nginx-deployment-59bdc8fccc-4pncp 0/1 Pending 0 8m10s

nginx-deployment-59bdc8fccc-6qb6p 0/1 Pending 0 8m10s

nginx-deployment-59bdc8fccc-8nscb 0/1 ContainerCreating 0 8m9s

nginx-deployment-59bdc8fccc-9cgwt 0/1 ContainerCreating 0 8m10s

nginx-deployment-59bdc8fccc-f7mfz 0/1 ContainerCreating 0 8m10s

nginx-deployment-59bdc8fccc-l6q7q 0/1 Pending 0 8m10s

nginx-deployment-59bdc8fccc-ljmfn 0/1 ContainerCreating 0 8m10s

nginx-deployment-59bdc8fccc-rpj9v 0/1 ContainerCreating 0 8m10s

vote-5d548444c5-5tgbf 1/1 Running 0 2d1h

vote-5d548444c5-rb6rz 1/1 Running 0 2d1h

vote-5d548444c5-rsg4x 1/1 Running 0 2d1h

vote-5d548444c5-s5kps 1/1 Running 0 2d1h

web-699689dbd-85kqx 0/1 ContainerCreating 0 8m10s

web-699689dbd-9p6w5 0/1 Pending 0 8m10s

web-699689dbd-k8nnz 0/1 Pending 0 8m10s

anuj@maste01:~$ kubectl get pod

NAME READY STATUS RESTARTS AGE

nginx-deployment-59bdc8fccc-4pncp 0/1 Pending 0 8m16s

nginx-deployment-59bdc8fccc-6qb6p 1/1 Running 0 8m16s

nginx-deployment-59bdc8fccc-8nscb 1/1 Running 0 8m15s

nginx-deployment-59bdc8fccc-9cgwt 1/1 Running 0 8m16s

nginx-deployment-59bdc8fccc-f7mfz 1/1 Running 0 8m16s

nginx-deployment-59bdc8fccc-l6q7q 1/1 Running 0 8m16s

nginx-deployment-59bdc8fccc-ljmfn 1/1 Running 0 8m16s

nginx-deployment-59bdc8fccc-rpj9v 1/1 Running 0 8m16s

vote-5d548444c5-5tgbf 1/1 Running 0 2d1h

vote-5d548444c5-rb6rz 1/1 Running 0 2d1h

vote-5d548444c5-rsg4x 1/1 Running 0 2d1h

vote-5d548444c5-s5kps 1/1 Running 0 2d1h

web-699689dbd-85kqx 0/1 ContainerCreating 0 8m16s

web-699689dbd-9p6w5 0/1 ContainerCreating 0 8m16s

web-699689dbd-k8nnz 1/1 Running 0 8m16s

anuj@maste01:~$

//* so by doing above type of secheduling we can controll pod scheduling if each and eveary application developer use nodeSelector or nodeName

but if any of the developer not use such controlls then, that application pod will deploy on any of the node.

so to over come thing we can use taints and tolleration *//

====================================================

Taint and Tolerations

=================================================

Now just for lab lets remove the taint from master node and put taint on worker node

anuj@maste01:~$ kubectl describe node/worker01 | grep -i taint

Taints: <none>

anuj@maste01:~$ kubectl taint node maste01 node-role.kubernetes.io/master-

node/maste01 untainted

anuj@maste01:~$ kubectl describe node/maste01 | grep -i taint

Taints: <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl taint node worker01 team=uat:NoSchedule

node/worker01 tainted

anuj@maste01:~$

anuj@maste01:~$ kubectl taint node worker02 team=sit:NoSchedule

node/worker02 tainted

/* Now lets create one deploeyment and see pod should not run on nodes those have taint applied */

anuj@maste01:~$ cat apache.yaml

apiVersion: apps/v1

kind: Deployment

metadata:

labels:

app: apache

namespace: default

spec:

progressDeadlineSeconds: 600

replicas: 3

revisionHistoryLimit: 10

selector:

matchLabels:

app: apache

strategy:

rollingUpdate:

maxSurge: 25%

maxUnavailable: 25%

type: RollingUpdate

template:

metadata:

creationTimestamp: null

labels:

app: apache

spec:

containers:

- image: httpd

imagePullPolicy: Always

anuj@maste01:~$

anuj@maste01:~$ kubectl apply -f apache.yaml

deployment.apps/apache created

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

apache-5667776978-97dqv 0/1 ContainerCreating 0 5s <none> maste01 <none> <none>

apache-5667776978-swpwh 0/1 ContainerCreating 0 5s <none> maste01 <none> <none>

apache-5667776978-z9r54 0/1 ContainerCreating 0 5s <none> maste01 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

apache-5667776978-97dqv 0/1 ContainerCreating 0 8s <none> maste01 <none> <none>

apache-5667776978-swpwh 0/1 ContainerCreating 0 8s <none> maste01 <none> <none>

apache-5667776978-z9r54 1/1 Running 0 8s 10.46.0.1 maste01 <none> <none>

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

apache-5667776978-97dqv 1/1 Running 0 11s 10.46.0.2 maste01 <none> <none>

apache-5667776978-swpwh 0/1 ContainerCreating 0 11s <none> maste01 <none> <none>

apache-5667776978-z9r54 1/1 Running 0 11s 10.46.0.1 maste01 <none> <none>

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

apache-5667776978-97dqv 1/1 Running 0 12s 10.46.0.2 maste01 <none> <none>

apache-5667776978-swpwh 1/1 Running 0 12s 10.46.0.3 maste01 <none> <none>

apache-5667776978-z9r54 1/1 Running 0 12s 10.46.0.1 maste01 <none> <none>

/* all the pods are scheduled on master node as there is no taint on it and on the worker node there is taint applied hence no pod schedule on it

untill we add pod tolleration that allow pod to schedule on tainted node */

now lets add pod tolleration so that can schedule the pod on tainted node.

anuj@maste01:~$ cat nginx-deployment.yaml

apiVersion: apps/v1

kind: Deployment

metadata:

labels:

app: nginx

spec:

strategy:

type: RollingUpdate

rollingUpdate:

maxSurge: 2

maxUnavailable: 0

revisionHistoryLimit: 4

replicas: 8

minReadySeconds: 20

selector:

matchLabels:

app: nginx

template:

metadata:

labels:

app: nginx

spec:

containers:

- name: nginx

image: nginx:1.14.2

ports:

- containerPort: 80

tolerations: <=========== under this section we are telling schedular to schedule the pod on nodes that has taint

- key: "team" team:sit:NoSchedule

operator: "Equal"

value: "sit"

effect: "NoSchedule"

anuj@maste01:~$

anuj@maste01:~$ kubectl describe node/worker01 | grep -i taint

Taints: team=uat:NoSchedule

anuj@maste01:~$

anuj@maste01:~$ kubectl describe node/worker02 | grep -i taint <======= worker02 has taint that is mentioned in deploeyment, so adding

Taints: team=sit:NoSchedule pod tolleration to node02 to allow to schedule pod on it.

anuj@maste01:~$

anuj@maste01:~$ kubectl describe node/maste01 | grep -i taint

Taints: node-role.kubernetes.io/master:NoSchedule

anuj@maste01:~$

//* Now lets apply the deploeyment and pod should be schedule on worker02 as for that we have added pod taint tolleration *//

anuj@maste01:~$ kubectl apply -f nginx-deployment.yaml

deployment.apps/nginx-deployment created

anuj@maste01:~$

//* Successfully assigned default/nginx-deployment to worker02 *//

anuj@maste01:~$ kubectl describe pod/nginx-deployment-5d458f89b5-j7dc2 | tail -5

---- ------ ---- ---- -------

Normal Scheduled 6m40s default-scheduler Successfully assigned default/nginx-deployment-5d458f89b5-j7dc2 to worker02

Normal Pulled 6m27s kubelet Container image "nginx:1.14.2" already present on machine

Normal Created 6m27s kubelet Created container nginx

Normal Started 6m26s kubelet Started container nginx

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-5d458f89b5-47t2z 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-5vvgq 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-b9lc2 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-fjjgw 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-ghl8g 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-h5r5z 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-j7dc2 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-tjlll 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-5d458f89b5-47t2z 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-5vvgq 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-b9lc2 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-fjjgw 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-ghl8g 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-h5r5z 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-j7dc2 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-tjlll 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-5d458f89b5-47t2z 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-5vvgq 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-b9lc2 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-fjjgw 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-ghl8g 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-h5r5z 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-j7dc2 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-tjlll 0/1 ContainerCreating 0 13s <none> worker02 <none> <none>

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-5d458f89b5-47t2z 0/1 ContainerCreating 0 16s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-5vvgq 0/1 ContainerCreating 0 16s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-b9lc2 1/1 Running 0 16s 10.40.0.1 worker02 <none> <none>

nginx-deployment-5d458f89b5-fjjgw 1/1 Running 0 16s 10.40.0.4 worker02 <none> <none>

nginx-deployment-5d458f89b5-ghl8g 1/1 Running 0 16s 10.40.0.2 worker02 <none> <none>

nginx-deployment-5d458f89b5-h5r5z 0/1 ContainerCreating 0 16s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-j7dc2 0/1 ContainerCreating 0 16s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-tjlll 0/1 ContainerCreating 0 16s <none> worker02 <none> <none>

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-5d458f89b5-47t2z 1/1 Running 0 22s 10.40.0.3 worker02 <none> <none>

nginx-deployment-5d458f89b5-5vvgq 1/1 Running 0 22s 10.40.0.7 worker02 <none> <none>

nginx-deployment-5d458f89b5-b9lc2 1/1 Running 0 22s 10.40.0.1 worker02 <none> <none>

nginx-deployment-5d458f89b5-fjjgw 1/1 Running 0 22s 10.40.0.4 worker02 <none> <none>

nginx-deployment-5d458f89b5-ghl8g 1/1 Running 0 22s 10.40.0.2 worker02 <none> <none>

nginx-deployment-5d458f89b5-h5r5z 1/1 Running 0 22s 10.40.0.10 worker02 <none> <none>

nginx-deployment-5d458f89b5-j7dc2 1/1 Running 0 22s 10.40.0.9 worker02 <none> <none>

nginx-deployment-5d458f89b5-tjlll 1/1 Running 0 22s 10.40.0.8 worker02 <none> <none>

anuj@maste01:~$

//* Now by adding pod tolleration we make taineted node as candidate to schedule the pod on the node that has taint, also the node that has

no taints applied also become candidate to schedule the pods, now lets see the practicale to prove this

as of now

-> we have taint on master and both worker node,

-> for the worker02 we have added tolleration, so this node is candidate to schedule the pod.

-> worker01 also have taint to team=sit

-> mastero1 also have noSechedule taint, so at this momment we dont have any node that has taint is not applied lets remove the taint from the

master node and you will see pod will schedule on master node as well as we have removed the taint and make it candidate to schedule the pods.

anuj@maste01:~$ kubectl describe node/maste01 | grep -i taint

Taints: node-role.kubernetes.io/master:NoSchedule

anuj@maste01:~$

anuj@maste01:~$ kubectl describe node/worker01 | grep -i taint

Taints: team=uat:NoSchedule

anuj@maste01:~$

anuj@maste01:~$ kubectl describe node/worker02 | grep -i taint

Taints: team=sit:NoSchedule

/* now lets remove taint from master node *//

anuj@maste01:~$ kubectl taint node maste01 node-role.kubernetes.io/master-

node/maste01 untainted

anuj@maste01:~$

anuj@maste01:~$anuj@maste01:~$ cat nginx-deployment.yaml

apiVersion: apps/v1

kind: Deployment

metadata:

labels:

app: nginx

spec:

strategy:

type: RollingUpdate

rollingUpdate:

maxSurge: 2

maxUnavailable: 0

revisionHistoryLimit: 4

replicas: 8

minReadySeconds: 20

selector:

matchLabels:

app: nginx

template:

metadata:

labels:

app: nginx

spec:

containers:

- name: nginx

image: nginx:1.14.2

ports:

- containerPort: 80

tolerations: <=========== we have pod tolleration that matches for worker02.

- key: "team"

operator: "Equal"

value: "sit"

effect: "NoSchedule"

anuj@maste01:~$

anuj@maste01:~$ kubectl apply -f nginx-deployment.yaml

deployment.apps/nginx-deployment created

anuj@maste01:~$

//* please refer following output you can see pods are scheduled on worker02 along with maste01 as well as we dont have any taint on master node.

hence this is a problem, as we want our pod should be schedule on node that has taint team=uat *//

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-5d458f89b5-7tb65 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-hsw2q 0/1 ContainerCreating 0 4s <none> maste01 <none> <none>

nginx-deployment-5d458f89b5-lfdhx 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-n8qw7 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-qqnn9 0/1 ContainerCreating 0 4s <none> maste01 <none> <none>

nginx-deployment-5d458f89b5-s8tww 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-sg9qf 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-zprm9 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl get pod -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-5d458f89b5-7tb65 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-hsw2q 1/1 Running 0 9s 10.46.0.2 maste01 <none> <none>

nginx-deployment-5d458f89b5-lfdhx 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-n8qw7 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-5d458f89b5-qqnn9 1/1 Running 0 9s 10.46.0.1 maste01 <none> <none>

nginx-deployment-5d458f89b5-s8tww 1/1 Running 0 9s 10.40.0.2 worker02 <none> <none>

nginx-deployment-5d458f89b5-sg9qf 1/1 Running 0 9s 10.40.0.1 worker02 <none> <none>

nginx-deployment-5d458f89b5-zprm9 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

anuj@maste01:~$

//* Now to over come the above situation we have to add the nodeSelector as well along with pod toleration in the deploeyment config *//

lets see all of the above node taint and label reamin same and using nodeSelector pods should be sechedule on worker02 only not on

any other node now. *//

anuj@maste01:~$ cat nginx-deployment.yaml

apiVersion: apps/v1

kind: Deployment

metadata:

labels:

app: nginx

spec:

strategy:

type: RollingUpdate

rollingUpdate:

maxSurge: 2

maxUnavailable: 0

revisionHistoryLimit: 4

replicas: 8

minReadySeconds: 20

selector:

matchLabels:

app: nginx

template:

metadata:

labels:

app: nginx

spec:

containers:

- name: nginx

image: nginx:1.14.2

ports:

- containerPort: 80

tolerations:

- key: "team"

operator: "Equal"

value: "sit"

effect: "NoSchedule"

nodeSelector:

team: sit

anuj@maste01:~$

anuj@maste01:~$ kubectl apply -f nginx-deployment.yaml

deployment.apps/nginx-deployment created

anuj@maste01:~$

//* Just refer the following output pods are scheduled on worker02 only this time, not on any other node, hence our problem is solved *//

anuj@maste01:~$ kubectl get po -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-fbcb7d9d9-4lj9m 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-j8pgr 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-lzrn7 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-mqljj 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-qcps4 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-rjfz5 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-zn4nl 0/1 ContainerCreating 0 4s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-znzbw 0/1 ContainerCreating 0 5s <none> worker02 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl get po -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-fbcb7d9d9-4lj9m 0/1 ContainerCreating 0 10s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-j8pgr 0/1 ContainerCreating 0 10s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-lzrn7 0/1 ContainerCreating 0 10s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-mqljj 0/1 ContainerCreating 0 10s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-qcps4 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-rjfz5 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-zn4nl 0/1 ContainerCreating 0 9s <none> worker02 <none> <none>

nginx-deployment-fbcb7d9d9-znzbw 1/1 Running 0 10s 10.40.0.2 worker02 <none> <none>

anuj@maste01:~$

anuj@maste01:~$ kubectl get po -o wide

NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES

nginx-deployment-fbcb7d9d9-4lj9m 1/1 Running 0 13s 10.40.0.9 worker02 <none> <none>

nginx-deployment-fbcb7d9d9-j8pgr 1/1 Running 0 13s 10.40.0.8 worker02 <none> <none>

nginx-deployment-fbcb7d9d9-lzrn7 1/1 Running 0 13s 10.40.0.7 worker02 <none> <none>

nginx-deployment-fbcb7d9d9-mqljj 1/1 Running 0 13s 10.40.0.1 worker02 <none> <none>

nginx-deployment-fbcb7d9d9-qcps4 1/1 Running 0 12s 10.40.0.4 worker02 <none> <none>

nginx-deployment-fbcb7d9d9-rjfz5 1/1 Running 0 12s 10.40.0.3 worker02 <none> <none>

nginx-deployment-fbcb7d9d9-zn4nl 1/1 Running 0 12s 10.40.0.10 worker02 <none> <none>

nginx-deployment-fbcb7d9d9-znzbw 1/1 Running 0 13s 10.40.0.2 worker02 <none> <none>

anuj@maste01:~$

Unix Server Solutions

Sunday, May 1, 2022

kubernetes Pod Scheduling

No comments:

Post a Comment

kubernetes Pod Scheduling

Search This Blog