Table of Contents

Docker

What is Docker? what is contaienr

What is a container

  • a way to package application with all the dependencies
  • portable artifact

The area where containers are stored Container Repository

  • private repo
  • public repo (dockerhub)
How container improved application development

for developing application individual services needs to be added in host machine which mighe be different for different applications.

after that we can have our

  • own isolated environment
  • consistent application development
  • applicaiton deployment

    developer

    • application binary with instructions, databases with instructions

    operations

    • install service with external dependencies
    • conflict in dependencies
    • miscommunication due to textual nature of instructions

    After Containers developer and operation work together to create docker env No configuration on server except docker runtime

What is container

  • layer of images
  • and base layer is linux (mostly alpine)

application layer (postgres)

Linux Base Image

Image Vs Container

artifact

running the image is container

Docker vs Virtual Machine

operating system has 2 components

  • application
  • kernel

virutal machine abstract both components application as well as kernel so they are intensive to run

container use the kernel of host operating system and that and only include application layer along with dependencies.

Docker installation

already done

Commands

contaienr Vs image

container = running image

container = filesystem, environemnt configs, application image, port

  • virtual file system
  • port: talk to appln. inside container
  • application image - redis

docker pull redis docker run redis

docker ps can be used to check the containers C -c to kill it

to start in detached mode docker run -d redis

if some application that we are running crashed we can stop that and start it again to do that

using docker ps command to get Container ID

docker stop <container id>

docker start <container id>

well remembering the container id might be a pain if we stopped it and started looking for it again after some time

docker ps -a will give all the container - stopped as well as running.

docker run will do contitional pull and start.

Binding container ports to host ports

when we run containers, it runs on a port and that port is container port and we cannot access it via that port

for example. redis runs on 6379 and that port, and if we run it again of different version the sample port will get used it is possible because its container's port and that's why it can run on same port. to attach it to a host's port what we can do is specify host port when we run a container.

i.e.

docker run -p<hostport>:<containerport> -d image:version 

docker run -p6000:6379 -d redis

docker run -p6000:6379 -d redis:4.0 # this will not work, conflict with host port

docker run -p6001:6379 -d redis:4.0 # this will work

Debugging a container

Specifying name while running container

docker run -d -p6001:6379 –name redislatest6001 redis

accessing the terminal

docker exec -it redislatest6000 /bin/bash

Developing with Container

Running MongoDB *docker run -d -p 27017:27017 -e MONGOINITDBROOTUSERNAME=admin -e MONGOINITDBROOTPASSWORD=password –name mongodb –net animemedianetwork mongo

Running MongoDB Express docker run -d \ -p 8081:8081 \ -e MECONFIGMONGODBADMINUSERNAME=admin \ -e MECONFIGMONGODBADMINPASSWORD=password \ –net animemedianetwork \ –name mongoexpress \ -e MECONFIGMONGODBSERVER=mongodb \ mongo-express

we can connect our node.js app to other docker contaiener like mongodb and build out the application

Docker compose - Running multiple Service

Running multiple containers cab be tedious at times when we have multiple docker containers with additional steps.

Docker compose lets us record the multi container setup through a yaml file

docker-compose comes preinstalled with docker.

docker-compose -f <yaml file> up | down 

version: 3.0
services:
  mongodb:
    image: mongo
    ports:
    - 27017:27017
    environment:
    - MONGO_INITDB_ROOT_USERNAME=admin
    - MONGO_INITDB_ROOT_PASSWORD=password
  mongo_express
    image: mongo-express
    ports:
    - 8081:8081
    environment:
    - ME_CONFIG_MONGODB_ADMINUSERNAME=admin
    - ME_CONFIG_MONGODB_ADMINPASSWORD=password
    - ME_CONFIG_MONGODB_SERVER=mongodb \

the network is automatically created.

Dockerfile

base image to start off

FROM <> # base image

ENV <>=<>

RUN <linux command>

COPY <hostlocation> <imagelocation>

CMD ["command", "file"] # only one cmd command which is entrypoint of server

WORKDIR <container location>

build docker image

docker build -t name:tag <location of docker file> .

deleting images

docker rmi <imageId>

docker rm <contaienrId>

private docker Repos

Steps to work login in the aws ecr docker login build image push image

Image naming concept registryDomain/image:tag so for pulling docker pull registryDomain/image:tag

Deploying the containerized group

IMPORTANT when using containerized app, we need to change the connection string of mongodb from ip address to the container name.

Volumes Demo

for persistence in containers

a directory in host FS is mounted in container FS

Volume Types

docker run

  1. host volume

-v hostdirectory:containerdirectory

  1. anonymous volume

-v contaienrdirectory internally host volume will be managed by docker /var/lib/docker/volumes/random-hash/data

  1. named volume

-v name:containerdirectory