Skip to content

Scheduler

sheduler

manual schedulin

every yml file pod definiton has nodename sheduler looks who doesnot have it and runs scheduling algorithm and binds pod to node

if there is not scheduler pods will be in a pending state

1
2
spec:
  nodeName: node02

we cann not update it after pod creation but we can update it with post request

Labels and selectors

Labels are properties attached to items

selectors help to filter labels

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
 apiVersion: v1
 kind: Pod
 metadata:
  name: simple-webapp
  labels:
    app: App1
    function: Front-end
 spec:
  containers:
  - name: simple-webapp
    image: simple-webapp
    ports:
    - containerPort: 8080

kubectl get pods --selector app=App1

to creaate replicaset with connected to pods

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
 apiVersion: apps/v1
 kind: ReplicaSet
 metadata:
   name: simple-webapp
   labels:
     app: App1
     function: Front-end
 spec:
  replicas: 3
  selector:
    matchLabels:
     app: App1
 template:
   metadata:
     labels:
       app: App1
       function: Front-end
   spec:
     containers:
     - name: simple-webapp
       image: simple-webapp

annotations

annotations: buildversion: 1.2 record other details for info

taint and tolerations

allow certain nodes to accept only specific pods

kubectl taint nodes node1 app=blue:Noschedule

node affinity

certain pods to exact node

in master taint automaticly is added

kubectl describe node kubemaster | grep Taint

Node selectors

add nodeSelector in pods spec size: Large

label nodes

kubectl label nodes nodename size=Large

limitation u can not small or medium

node affinity

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
apiVersion: v1
kind: Pod
metadata:
 name: myapp-pod
spec:
 containers:
 - name: data-processor
   image: data-processor
 affinity:
   nodeAffinity:
     requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: size
            operator: In
            values: 
            - Large
            - Medium

 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
apiVersion: v1
kind: Pod
metadata:
 name: myapp-pod
spec:
 containers:
 - name: data-processor
   image: data-processor
 affinity:
   nodeAffinity:
     requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: size
            operator: NotIn
            values: 
            - Small
            ```

## requiredDuringSchedulingIgnoredDuringExecution 

only this type of node

## prefferedDuringSchedulingIgnoredDuringExecution 
if not found sheduler ignore affinity rules


## requiredDuringSchedulingRequirdDuringExecution
bad pods automaticly will be deleted


## taint and toleration together
to place exacly where we want

## Daemon Sets

for example we want logging agent in all cluster and node
or kube-proxy can be deployed as daemon set

```yml
apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: monitoring-daemon
  labels:
    app: nginx
spec:
  selector:
    matchLabels:
      app: monitoring-agent
  template:
    metadata:
     labels:
       app: monitoring-agent
    spec:
      containers:
      - name: monitoring-agent
        image: monitoring-agent

kubectl get daemonset kubectl describe daemonset

deemon sets uses node affinity and taint after 1.x verion before was nodeName:

what we want custom sheduling program.

kubernetes cluster can have multiple shedulers

service options: --scheduler-name:custom-scheduler

or name int /etc/kubernetes/manifest/kube-sheduler/yaml

and add in command: --scheduler-name=custom-scheduler

if there is multiple replicas sheduler only one is active

there is election process who will be leader

where is parameter to aviod newly created schedulers to get leaders

--lock-object-name=custom-scheduler

get logs

kubectl logs custom-scheduler --name-space=kube-system

add taint node

kubectl taint nodes nodes01 spray=mortein:NoSchedule

remove taint node

kubectl taint nodes controlplane node-role.kubernetes.io/master:NoSchedule-

add label to node

kubectl label node node01 color=blue

get all daemonsets

kubectl get daemonsets --all-namespaces

describe daemon shedulet pods

kubectl describe daemonset kube-proxy --namespace=kube-system

create daemonset yaml

kubectl create deployment elasticsearch --image=k8s.gcr.io/fluentd-elasticsearch:1.20 -n kube-system --dry-run=client -o yaml > fluentd.yaml

create addidional scheduler from file

kubectl create -n kube-system configmap my-scheduler-config --from-file=/root/my-scheduler-config.yaml