/kubernetes

kubernetes

terms used in K8

important commands

kubectl get nodes,namespaces,services,deployments,replicaset,pods,secrets,configMaps -o wide

kubectl create deployment

kubectl apply -f configFileName --namespace=namespaceName

kubectl describe pods

kubectl logs podname

kubectl exec -it podnme command

Why Namespaces

Consider this like a portlets within Portal container where every portet has its own portlet-id . Every portlet has its own boundry but at the same time ,a portlet is allowed to interact with global objects. Also other portlets can access this portet with its portlet id. (but this statement is very generic and read the below to get more and more details)

  1. can be used to group similar applications
  2. Conflicts : Many teams , same application
  3. Resource Sharing : Staging and Development aka different environments
  4. Access and Resource Limits on Namespaces

important note

  1. you cant access most resources from another Namespaces example of resources which cannot be shared are but not limited to configMap, secrets Resources which can be shared are services. example for excessing a service deployed in another namespace ex application in namespace 1 want to access database service in another namespace then the application in namespace 1 should refer like serviceName.namespace

  2. There are certain components which cant be created in namespace i.e. which are global cluster,cant isolate them example volumes,nodek

    kubectl api-resources --namespaced=false (resources which are not namespaces aka global ) kubectl api-resources --namespaced=true (resources which can stay in boundaries of namespaces or can be confined in namespaces)

    when components are created in a cluster without a namespace ,components are created in default namespace

K8 Ingress

to do

Helm

Package manager for k8 like npm, brew,apt. It packages YAML files and distributes and distributes them in public and private repositories

Helm charts

Bundle of YAML files Create your own Helm Charts with Helm Push them to Heml Repository

Common components like DB, redis,elk etc are commonly available so can be reused and this thr are publiv and private registeries for helm charts. It is also a templating engine in which references can be defined in YAML files instead of actual values like

name : {{ .Values.container.name }} where this Values object is referenced from a values.yaml file. Values object is either defined via YAML file or with --set flag in command line

Helm chart Structure : Templating Feture

myChart/ Chart.yaml values.yaml charts/ templates/ ... heml install <chartname> helm install --values=my-values.yaml <chartname> Important Note : Default values are picked up from values.yaml file but if you create a new yaml file lets say my-values.yaml and overide a value in this file, so overidden value will be picked from my-values.yaml file and remaining values will be picked from Default values.yaml file

Release Management

heml v.2 Uses a helm client(cli) and server(Tiller deployed in k8 cluster) to execute release management. Here Tiller stores the last state of configuration and we can use commands like

helm install helm upgrade helm rollback

Problem with this architecture

  • Tiller has too much power inside k8 cluster
  • Security issue

So in heml 3 , it was removed

K8 volumes