Docker 🐬 and Kubernetes 🚢
When writing this I am on MacOS with an M3 chip. and Rancher Desktop! But should work for everyone similarly.
brew install --cask rancher # This should install rancher desktop
In ⛭ Settings and preference I have
Container Engine : dockerd (moby)
Virtual Machine : VZ and VZ (Enable Rosetta support) checked
The first step is to Build an Image and Navigate the docker commands
Docker command structure is docker <type> <command>
For example: if you want to have any interaction with containers you will say
docker container <command>command can be as simple aslsto list containers
docker network <command> Similarly, if u want to do something with the network
#First things first create a Dockerfile
#Create a docker image by building it
docker build -t <image-name>:<tag> .
docker image ls #This will show all images
docker run --name my-container my-image:latest
docker container attach <"container_id"> #attach to container running in detach mode if u have opned it in -it
# This commands can also be used for network ie: docker network kill <"network_id">
docker container kill <"container_id"> #kills an running container immediately
docker container stop <"container_id"> #Gracefully stop container ie: 10s to stop resources and then kill
docker container prune # Remove all unused container
Additional tags for run command
#Additional Flags
#when you launch docker your microservice port could be internal to the container ie: 8080
#you want to expose this so will do -p 80:8080 so will be able to access application at http://yourwebiste.com:80
-p <local_machine_port>:<docker_port>
#This will print whats going inside container to terminal
--attach STDOUT --attach STDERR
#will remove or delete container after it exits
--rm
#Attach a volume (local)
-v "$(pwd)/local_folder":container_folder
#set working directory
-w /<working_directory>
#Run in detach mode means just show container id and exit
-d
#Open on interactive mode ie: provide STDInput in with pseudo TTL
-it
Connect to the container
# Use following command to connect to container when in bash
docker exec -it <container_id> /bin/bash
#Copy files & Folders in running container
docker cp <source_location> <container_id>:<desination_path>
Docker Networking 📡
Docker networks are of the following types (DRIVER)
None
default Bridge
User-defined Bridge
Host Network
MacVLAN Network
Overlay Network
IPvLan
#Here are basic commands that we would use often for networking
docker network ls # list networks
docker network inspect <network_name> #
docker network connect <network_id> <container_id>
docker network disconnect <network_id> <container_id>
#1. <default-Bridge>
#This is the default the network driver when you create a container
#It shows as interface:docker0 and gateway:172.17.0.1 subnet:172.17.0.0/16
docker container run --name "nginx" -d nginx:latest
#this will create a nginx container and attach it to default network bridge
#Default bridge creates a virtual network interface, this can be checked by `ifconfig` will see something like `vethe879@if1`
#And acts as DHCP so assignes a IP address, this can be checked by docker network inspect command
#Means ! - By default
# Containers reach Internet
# You cannot reach Container | Expose port to achieve this -p 80:80
# All container within bridge can communicate with each other
#2. <User-defined-Bridge>
docker network create <"network-name">
#This will create a new network, Below is example how you attach container to network
docker container run --network "network-name" --name "nginx" -d nginx:latest
#Means ! - By default
# Container in other networks cannot reach container in network-name
# Should also get DNS meaning ping <"container-name"> should work ( Did not work for me on rancher desktop ))
#3. <host>
docker container run --network host --name "nginx" -d nginx:latest
#This will run application with same ip as host, just like running application installed on your machine
#4. <MacVLAN>
#MacVLAN allow us to connect the docker container directly to our physical network
#They will get their own MAC Address and IP address on network ie: Home network
#ToDo: READ on Vlan
docker network create -d macvlan \
--subnet <YOUR_SUBNET> \
--gateway <YOUR_GATEWAY> \
-o parent=<YOUR_PARENT_INTERFACE> \
<NETWORK_NAME>
#1.Subnet : Home network subnet 2.Gateway : Routers Address
#3.O Flag : VLAN interface of the system where container will be deployed ie: eth0.10 this can be determined by using ifconfig
#Creating container in network
docker container run --network "network_name" --ip "10.7.1.22" --name "nginx" -d nginx:latest
#--ip this should be unique IP address in network
# NOTE: For this to work you might need to Turn ON promiscuous mode on each network device.
# WARNING: If you dont specify --ip docker dhcp will assign random address which could conflict with network dhcp
# INFO: MacVLAN has 2 modes above is (bridge mode) another mode is (802./q)
#4. <IPVLAN> (L2)
#Allows Host to share MacAddress with container (which solves promiscuous issue existed in macvlan )
#So each container will have unique IP but same macaddress as Host
#Router/Switch in the network should be Okay with this otherwise there is a problem
docker network create -d ipvlan \
--subnet <YOUR_SUBNET> \
--gateway <YOUR_GATEWAY> \
-o parent=<YOUR_PARENT_INTERFACE> \
<NETWORK_NAME>
#5. <IPVLAN> (L3)
#This connects containers directly to Hosts network interface. which solves all the problems
#But network does not know where to reach container and there is no internet connectivity
#To Achieve network you have to configure router table and tell it
#ToDO: Skipping this for Now
#6 <Overlay network>
#This is used in case of multiple host (used in kubernaties / Docker Swarm)
#7 <None network>
#Means no connectivity !
docker container run --network None --name "nginx" -d nginx:latest
Connect to eks using kubectl
Setup was profile locally for this method
log in using `aws sso login --profile <your-profile>`
For more Info on AWS commands and setup check aws blog
aws eks update-kubeconfig --name <cluster-name> \
--region <region> \
--role-arn <role that gives access to eks> \
--profile <aws saved profile> \
--alias <context_name>
kubectl config get-contexts #This will list all kube context
kubectl config use-context <context_name> #this will select the context name
kubectl config delete-context <context-name> #Will delete the context
Kubectl commands
# Namespace
kubectl get namespace #This will show all the namespace available
kubectl create namespace <namespace-name> #Creates new namespace
kubectl delete namespace <namespace-name> #Delete's namespace
#Pod
kubectl get pod -n <namespace>
kubectl delete pod <pod-name> -n <namespace>
kubectl logs <pod-name> -n <namespace>
#Ingress
kubectl get ingress -n <namespace>
kubectl describe ingress <ingress-name> -n <namespace>
#Services
kubectl get svc -n <namespace> #gives services in the namespace
kubectl describe svc -n <svc-name> -n <namespace>
Adding and checking secrets in eks
#Secrets
kubectl create secret generic <secret-name> \
--from-literal='key1=value1' \
--from-literal='key2=value2' \
--namespace=<namespace-name> # This will create a secret <secret-name> with key:value pairs
#check if secret has been created
kubectl describe secret <store-name> --n <namespace-name>
#This will give json output, the values will be base64 encoded
kubectl get secret <store-name> --n <namespace-name> -o json
#Connect to a POD
kubectl exec -it <pod-name> -n <namespace> -- /bin/bash
dget pods
create namespace
Commands//
what is ssh config file