e2e-hello CASE Documentation

This Project contains couple of major point shown below.

• Application

  • Codebase
  • Helm Chart

• Kubernetes Cluster (gcloud)

• Monitoring

  • Prometheus/Grafana
  • PrometheusAdapter/CustomMetrics
  • PrometheusExporter of application

• CI/CD Pipeline

  • • Jenkins
  • • Docker/Helm

Application

CodeBase

This is a simple python application you can see the codebase and Dockerfile under application directory. You can build shown on docker environment.

    $ docker build .

Local:

    $ pip3 install -r requirements.txt
    $ python3 app.py

Application contains prometheus exporter and it returning /metrics endpoint successfully to gather metrics./hi endpoint returins valid JSON response.

HelmChart

HelmChart located on application/hello-world directory.You can update values under values.yaml file.

KUBERNETES CLUSTER

I have setup a cluster on google-cloud for development purposes. The command is shown as below:

    $ gcloud container clusters create MY_CLUSTER \
        --zone compute-zone

Notice: Metric-server already installed on GKE based clusters.If they don't I prefer to use this manifest file shown below: NginxIngress able to us accept traffic easily:

    $ helm install --name nginx-ingress stable/nginx-ingress

Monitoring

You can see general architectural diagram with explanation shown below. I have used prometheusoperator with helm chart, it has couple of ability to access Dasboard and a lots of CRD options to easily automate and manage prometheus via simple manifest files.

In this project you can see the installation scripts under ./create_cluster/components.sh directory. By the I have added an example to get alert activity via webhook please check the directory autoscaler directory.It is a simple application that listen incoming payload from Alert Manager, to send the alerts via webhook you see the definition under cluster operator.yaml file.

cluster-autoscaler-rules containing application level alert defintion for prometheus and alertmanager, when you set up operator those definiton will appear on prometheus monitor.

CI/CD Pipeline

I have used jenkins to build and deploy application.I set up jenkins via helm chart, under the jenkins directory, it is has a installation script setup_jenkins.sh

Jobs are configured to poll changes from github.Jenkins configured to use kubernetes for run the runner agent to execute the jobs.

You can see the diagram below.