CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
Prerequisites:
General Requirements:
1. Microsoft Visual Studio Code – Download Visual Studio Code - Mac, Linux, Windows
2. AWS Account with Programmatic Access (Access and Secret Key) – Watch the Video for steps
3. Python Installation - Download Python | [Link]
4. Docker Installation - Install Docker Engine | Docker Documentation
NB: The general requirements apply to all Operating Systems
Windows Users:
1. Install Microsoft C++ Build Tools - Microsoft C++ Build Tools - Visual Studio
2. Enable Virtualization in your BIOS Settings (Needed to run Docker Desktop)
3. RUN Docker Desktop Using WSL (Windows Subsystem for Linux)
4. Open Powershell and run wsl --update
5. Then you can launch your Docker Desktop and Start the VSCode Editor.
PYTHON APP DEPLOYMENT ON AWS EKS
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
STEP 1
To start, create a new folder and open it using VSCode. Create a new file with the name [Link]
and paste the code below-
Flask
MarkupSafe
Werkzeug
itsdangerous
psutil
plotly
tenacity
boto3
kubernetes
Before running the above script, check the following (Windows User):
python --version
Python 3.10.7
pip --version
pip 23.1 from C:\Users\hp\AppData\Local\Programs\Python\Python310\lib\site-packages\pip
(python 3.10)
Before running the above script, check the following (Linux and Mac Users):
Python3 --version
Python 3.10.6
pip --version
pip 22.0.2 from /usr/lib/python3/dist-packages/pip (python 3.10)
If the above results are returned, then install the requirements file using the following commands:
Windows:
pip install -r .\[Link]
Linux and Mac
pip3 install -r [Link]
STEP 2
Authenticate with your AWS Account using your access and secret key; run the following commands:
aws configure
AWS Access Key ID [****************WZHI]: #Replace with your access key
AWS Secret Access Key [****************T+ys]: #Replace with your secret key
Default region name [eu-west-2]: #Choose your own region
Default output format [None]: #You can leave this as none
NB: Ensure your IAM Account as an AdministratorsAccess Policy attached. Do not use your root
account.
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
Then create a new file within your VSCode named [Link] (to create an Elastic Container Registry for
our Docker Image). Copy the script below and run it using python [Link](Windows) or python3(Linux
and MacOS) [Link] in your VSCode terminal
import boto3
client = [Link]('ecr')
response = client.create_repository(
repositoryName='flask-monitoring-app', # Any name of your choice
tags=[
{
'Key': 'flask-app-docker_image', # Any name of your choice
'Value': 'flask-monitoring-app' # Any name of your choice
},
],
imageScanningConfiguration={
'scanOnPush': True
}
)
ecr_uri = response['repository']['repositoryUri']
print(f"ECR URI=: {ecr_uri}")
STEP 3
Create a new file named [Link] copy and run the script below locally to test your application:
import psutil
from flask import Flask, render_template
app = Flask(__name__)
@[Link]("/")
def index():
cpu_metric = psutil.cpu_percent()
mem_metric = psutil.virtual_memory().percent
Message = None
if cpu_metric > 80 or mem_metric > 80:
Message = "High CPU or Memory Detected, scale up!!!"
return render_template("[Link]", cpu_metric=cpu_metric, mem_metric=mem_metric,
message=Message)
if __name__=='__main__':
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
[Link](debug=True, host = '[Link]')
If you have this page, Congratulations!
STEP 4
Package application code into a Docker Image using a Dockerfile. Create a new file named Dockerfile
and copy the below and paste.
# Use the official Python image as the base image
FROM python:3.11.3-buster
# Set the working directory in the container
WORKDIR /app
# Copy the requirements file to the working directory
COPY [Link] .
# Install the required Python packages
RUN pip3 install --no-cache-dir -r [Link]
# Copy the application code to the working directory
COPY . .
# Set the environment variables for the flask app
ENV FLASK_RUN_HOST=[Link]
# Expose the port on which the Flask app will run
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
EXPOSE 5000
# Start the Flask app when the container is run
CMD ["flask", "run"]
Use the following command to build a Docker image:
docker build -t flaskapp .
Check the new image using the following command:
docker images
To run the docker image and confirm if the app is running as it was locally, run the following
command:
docker run -d -p5000:5000 flaskapp:latest
You should have the following result:
Push the Docker image into the ECR repo created in Step 2. On your ECR page, click on the repo
and view push commands; it shows you steps to tag your docker image and push it to the
repository. You can watch the live demo video for more clarity.
STEP 5
Deploy the application on AWS EKS Cluster.
First, you need to create two IAM Roles.
1. Amazon EKS Cluster IAM Role
[Link]
2. Amazon EKS Node IAM Role
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
[Link]
Once you have the two roles created, create your EKS Cluster; watch the live demo video for a step-by-
step guide. To connect to your EKS cluster, run the following command to check that you installed
kubectl correctly.
Note: Before deploying your Kubernetes cluster, ensure your VPC and subnets are appropriately
configured. Also, Create a security group and add an incoming rule allowing port 5000.
kubectl version
================
Output
=========
WARNING: This version information is deprecated and will be replaced with the output from
kubectl version --short. Use --output=yaml|json to get the full version.
Client Version: [Link]{Major:"1", Minor:"25", GitVersion:"v1.25.4",
GitCommit:"872a965c6c6526caa949f0c6ac028ef7aff3fb78", GitTreeState:"clean", BuildDate:"2022-
11-09T[Link]Z", GoVersion:"go1.19.3", Compiler:"gc", Platform:"windows/amd64"}
Kustomize Version: v4.5.7
Unable to connect to the server: dial tcp: lookup [Link]-
[Link]: no such host
If you have the output above, then run the following command to connect your kubectl client to the
AWS EKS cluster:
aws eks update-kubeconfig --name flask-app-kube-cluster # Replace with the name of your
cluster
Then run the command below again; your output should be different from the previous
kubectl version
Create your Kubernetes deployment and service using a Python script. First, create a file named [Link],
and copy and paste the script below, then run python [Link] (Windows) or python3 [Link] (Linux and
Mac)
#create deployment and service
from kubernetes import client, config
# Load Kubernetes configuration
config.load_kube_config()
# Create a Kubernetes API client
api_client = [Link]()
# Define the deployment
deployment = client.V1Deployment(
metadata=client.V1ObjectMeta(name="my-flaskapp"), #Change the name
spec=client.V1DeploymentSpec(
replicas=1,
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
selector=client.V1LabelSelector(
match_labels={"app": "my-flaskapp"} #name must match metadata name above
),
template=client.V1PodTemplateSpec(
metadata=client.V1ObjectMeta(
labels={"app": "my-flaskapp"} #name must match metadata name above
),
spec=client.V1PodSpec(
containers=[
client.V1Container(
name="my-flaskcontainer", #name of your container
image="[Link]/flask-app-ecr-
repo", #URI of your docker image in ecr
ports=[client.V1ContainerPort(container_port=5000)]
)
]
)
)
)
)
# Create the deployment
api_instance = client.AppsV1Api(api_client)
api_instance.create_namespaced_deployment(
namespace="default",
body=deployment
)
# Define the service
service = client.V1Service(
metadata=client.V1ObjectMeta(name="my-flask-service"), #choose your own name
spec=client.V1ServiceSpec(
selector={"app": "my-flask-app"}, #choose your own name
ports=[client.V1ServicePort(port=5000)]
)
)
# Create the service
api_instance = client.CoreV1Api(api_client)
api_instance.create_namespaced_service(
namespace="default",
body=service
)
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
CLOUD-NATIVE PYTHON APP ON AWS EKS
TECHNICAL GUIDE
Run the following commands to check your deployments, pods, and service:
kubectl get deployments -n default
kubectl get pods -n default
kubectl get svc -n default
Deployments – To check if your container ran successfully
Pods – To check if the pods were successfully deployed on your nodes
SVC – To check that your container port is open to network communication.
To communicate with your app on AWS EKS locally, run the following command:
kubectl port-forward svc/<nameofservice> <port:port>
The above command lets you communicate with your EKS app using your localhost.
NB: When you run kubectl get svc -n default – you get the service name and port number
Note: The [Link] script above allows you to interact with your app locally. To expose your app
to the public (internet access), you need to configure the service type to use a loadbalancer. To do
that replace the Define the service section in the [Link] code with these:
# Define the service
service = client.V1Service(
metadata=client.V1ObjectMeta(name="my-flaskservice"),
spec=client.V1ServiceSpec(
selector={"app": "my-flaskapp"},
type="LoadBalancer",
ports=[client.V1ServicePort(
port=5000,
target_port=5000)]
)
)
NB: Before running the edited [Link] ensure you delete your initial deployment and service; use
the following commands:
kubectl delete deployments -n default
kubectl delete svc <servicename> -n default
When a deployment is deleted, all associated pods are deleted automatically. You can verify the
deletion by using the following commands:
kubectl get deployments -n default
kubectl get pods -n default
kubectl get svc -n default
If all is clear, rerun python [Link] and check the status of your deployment. To get the
loadbalancer URL use the command below:
kubectl get svc -n default
Website: h ps://[Link]; Email: training@[Link]; Telegram: h ps://[Link]/+rPFyH9NzQ9c3YzJk;
