Pod is the basic unit of work It is application or service Pod is an atomic unit of scheduling in Node Kubernetes manages pod and not container Kubernetes job is to keep the pods running based on desired state State is the pod up and running Health is the application in the pod running Pod shares the same resource with all containers in it Pod shares the same IP to all containers in it Preferred approach is one container in

a) Kubelete Main Kubernetes agent Registers node with cluster (decide who is the master node) Watches api server Instantiates the pods ( not creating the pod) Reports back to master Expose endpoint on 10255 b) kube-proxy Manages the kubernetes networking Assigning pod IP address (indirectly) All containers in a pod share a single IP Load balances across all pods in service Can use different network plugins/drivers like calico (CNI plugins) c) Container Engine Container Management Pulling images start/stop the containers

a) kube-apiserver Server of Api’s Front end to the Master Node Any request from Inside or outside the kubernetes is received by API server Exposes the REST API Consumes JSON (via manifest file) b) Cluser store Persist storage for cluster state and config Uses etcd to manage c) Kube-Controller-Manager Manager/Controller of Controllers such as Node Controller, Endpoint controller, Namespace controller etc Watches the changes in API Server Helps maintain the desired state d) Kube-scheduler Watches apiserver for new pods Schedule

Kubernetes manages 1 to many containers Load Balancing – Manages the load balance of multiple containers running with same port in one host Auto healing Keep desired state of containers Allow to view the data center as a computer by managing all containers hosted in multiple host Container orchestration Automated Rollbacks Auto Scaling – addressing scalability challenges Absorb changes quickly Workload placements Infrastructure abstraction

Docker supports several storage drivers, using a pluggable architecture. The storage driver controls how images and containers are stored and managed on your Docker host overlay2 is the preferred storage driver for all currently supported Linux distributions, and requires no extra configuration. The legacy overlay driver was used for kernels that did not support the “multiple-lowerdir” feature required for overlay2 All currently supported Linux distributions now provide support for this, and it is therefore deprecated. The aufs storage driver Was the preferred storage driver for Docker 18.06

docker file with 2 CMD, 2 ENTRYPOINT docker file with 1 CMD, 1 ENTRYPOINT

https://devopsschool.com/tutorial/docker/dockerfile/dockerfile-example-sample-lab.html Docker file Example and Sample Program – 1 Docker file Example and Sample Program – 2 Docker file Example and Sample Program – 3 Docker file Example and Sample Program – 4 Docker file Example and Sample Program – 5

FROM – base image to start creating the new image RUN – Run command in a container created from previous line/layer ARG – Set a variable for docker file execution and reuse using $(varname) ENV – Meta layer to set environment for the container EXPOSE – Sepcify information port for the container and also can be used during port forwarding COPY – Copy the file in same location as Docker file to container locationADD – Copy the file in same

root@ip-172-31-28-155:/var/lib/docker/overlay2# docker run -itd –name anoop_centos_2 centos460aebb7de8bde42112d32e5e37df7617fb8e0588a9bdccf79f47d98357cede7 [root@ip-172-31-28-155:/var/lib/docker/overlay2# docker ps | grep anoop460aebb7de8b centos “/bin/bash” 4 seconds ago Up 3 seconds anoop_centos_2 root@ip-172-31-28-155:/var/lib/docker/overlay2# docker attach 460aebb7de8b [root@460aebb7de8b /]# yum update [root@460aebb7de8b /]# yum install git [root@460aebb7de8b /]# yum install epel-release [root@460aebb7de8b /]# yum install -y ntp –> failed [root@460aebb7de8b /]# yum install httpd [root@460aebb7de8b /]# read escape sequenceroot@ip-172-31-28-155:/var/lib/docker/overlay2# docker ps | grep anoop460aebb7de8b centos “/bin/bash” 12 minutes ago Up 12 minutes anoop_centos_2 root@ip-172-31-28-155:/var/lib/docker/overlay2# docker commit -a “anoop” -m “centos with

root@ip-172-31-28-155:/var/lib/docker# docker run -itd –name anoop_ubuntu_base ubuntubcba3eb8cfb3d221c9f8f753a65c86ee5ea924d8c69438342d3e83d429d572a9root@ip-172-31-28-155:/var/lib/docker# docker ps | grep anoopbcba3eb8cfb3 ubuntu “bash” 5 seconds ago Up 4 seconds anoop_ubuntu_baseroot@ip-172-31-28-155:/var/lib/docker# docker attach bcba3eb8cfb3 root@bcba3eb8cfb3:/# apt-get update root@bcba3eb8cfb3:/# apt-get install git root@bcba3eb8cfb3:/# apt-get install httpd root@bcba3eb8cfb3:/# apt-get install mini-httpd root@bcba3eb8cfb3:/# which git/usr/bin/git root@bcba3eb8cfb3:/# mini_httpdbind: Address already in use root@bcba3eb8cfb3:/# apt-get install ntpReading package lists… Done root@ip-172-31-28-155:/var/lib/docker# docker ps | grep anoopbcba3eb8cfb3 ubuntu “bash” 6 minutes ago Up 6 minutes anoop_ubuntu_base root@ip-172-31-28-155:/var/lib/docker# docker commit -a “anoop” -m “ubuntu with git, ntp

Docker image is a collection of file system (Root FS, USER FS and APP FS) One copy of Docker image get attached to each container created from the image We can create multiple containers from same image docker image made up of file system layered over each other i.e Layer-n(image n)….. –> Layer1(image 1) –> Layer 0 (image 0) Layer 0 in the docker image is base image and requires root file system Higher layers always win when there is

The docker update command dynamically updates container configuration like cpu, memory etc. To specify more than one container, provide space-separated list of container names or IDs.

1. rabbitmq – https://hub.docker.com/_/rabbitmq 2. nginx – https://hub.docker.com/_/nginx 3. fluentd – https://hub.docker.com/_/fluentd 4. prometheus – https://hub.docker.com/r/ubuntu/prometheus 5. kibana – https://hub.docker.com/_/kibana

docker wait will wait for the containers to stop and then print the exit code from the container

Docker image: Docker image is a collection of file system (Root FS, User FS and App FS). Use “docker images” command to see the docker images Docker Container: Docker container is the runnable instance of docker image. Docker container contains 1 ROOT FS, 1 PID, 1 NET and 1 USER and attach ROOT FS, PID and NET to user. One copy of docker image get attached to each container Use “docker ps” command to see all docker containers with status

docker stop stops the container gracefully where as docker kill terminates the process immediately. docker stop showing status as “Exited (0)” and docker kill showing status as “Exited (137)”

docker pause suspend all process in the container. docker unpause unsuspend all process in the container