istio-k8s

Istio Configuration with Sample K8s Application

Istio Setup with Application

Run the below script, it will help to install istion on your cluster

$ sudo bash istio.sh

Istio Installation

Download Istio

First download the istio on your machine with below command

$ curl -L https://istio.io/downloadIstio | ISTIO_VERSION=1.7.1 TARGET_ARCH=x86_64 sh - 

Add the istioctl client to your path (Linux or macOS):

export PATH=$PWD/istio-1.7.1/bin:$PATH

Install Istio

Run the below command to install the istio on your cluster with demo profile NOTE: Demo Profile for testing, but there are other profiles for production or performance testing.

$ istioctl install --set profile=demo

Add a namespace label to instruct Istio to automatically inject Envoy sidecar proxies when you deploy your application later:

$ kubectl label namespace default istio-injection=enabled

Deploy the Application

Python web application architecture

Run the below command to install the python web application on your kubernetes cluster

$ git clone https://github.com/angudadevops/istio-k8s.git

$ kubectl apply -f deployment.yaml

Above command will deploy the python web application wiht its service

First install the gateway that will help to connect istio ingress

$ kubectl apply -f gateway/python-gateway.yaml

Now install the destination rule to map the specific version as a subest

$ kubectl apply -f destination-rule/python-destination-rule.yaml

Deploy Istio Rules

Now it's time to descide what type of virtual service type you want to use, below are the available methods

• HTTP Based Routing
• Weight Based routing
• Fault Injection Routing

1. HTTP Based Routing

Run the below command to install Http based routing with prefix and rewrite rules

$ kubectl apply -f virtual-service/python-http-prefix.yaml

2. Weight Based Routing

Run the below command to install weight based routing, this method is for Blue Green or Canary based deployments

$ kubectl apply -f virtual-service/python-weight-based.yaml

3. Fault Injection Routing

It's best practice to add fault injection for any application, as this will help you to wait and reroute the request to different prefix.

$ kubectl apply -f virtual-service/python-fault-injection.yaml

Access the Application

First get the istion-ingress gateway node port to access the application

Run the below command to check the node port of the ingress gateway, you should see something like below

$ kubectl get svc -n istio-system | grep istio-ingressgateway
istio-ingressgateway        NodePort    10.110.12.112   <none>        15021:32029/TCP,80:30144/TCP,443:32186/TCP,31400:30469/TCP,15443:30433/TCP

Access the web application with your host IP and Node Port

http://<host-IP>:<node-Port>/

1. HTTP Based Routing

Access the application with above url with context path as either /v1 or /v2. You should see the difference

2. Weight Based Routing

Access the application with context path as /app, you should see weight is distrubuted 50/50 to v1/v2 app versions

3. Fault Injection Routing

Access the application without context path, then it will reroute the request to v2 version application