OpenShift - Python - Ansible

Setting up Openshift cluster for local Dev/Test environment, explore the capabilities of Openshift by deploying sample python web application and test the deployed web application using ansible

These instructions will get you a OpenShift up and running on your local machine for development and testing purposes.

• python

• Flask

• Centos7

• OpenShift

• OKD

• Docker

• Ansible

This solution is setup and tested on Centos7

CentOS7

Python3

The following commands will set up Openshift on CentOS7 using OC client tool. Login into the CentOS7 machine and run the commands as root user. This installs Ansible, Docker 1.13 and OC 3.11.

yum update -y

rpm --import "https://sks-keyservers.net/pks/lookup?op=get&search=0xee6d536cf7dc86e2d7d56f59a178ac6c6238f52e"

yum install -y yum-utils wget git ansible

yum-config-manager --add-repo https://packages.docker.com/1.13/yum/repo/main/centos/7

yum makecache fast

yum install -y docker-engine-1.13.1.cs9-1.el7.centos

systemctl enable docker.service

systemctl start docker.service

wget https://github.com/openshift/origin/releases/download/v3.11.0/openshift-origin-client-tools-v3.11.0-0cbc58b-linux-64bit.tar.gz

tar -xf openshift-origin-client-tools-v3.11.0-0cbc58b-linux-64bit.tar.gz -C /usr/local/bin/ --strip-components=1

# Allow user to run docker commands
sudo usermod -aG docker $USER

edit the /etc/docker/daemon.json file and add the following:

    {

        "insecure-registries": [

            "172.30.0.0/16"

        ]

    }

After editing the config, reload systemd and restart the Docker daemon.

sudo systemctl daemon-reload

sudo systemctl restart docker

Run oc cluster up to run OpenShift cluster with public host name or public ip. This will take sometime to download and install Openshift

# Run openshift cluster
oc cluster up --public-hostname=$HOSTNAME/$IP 

While running the above command, OpenShift web UI redirects the console to https://127.0.0.1:8443 by default. This is the known issue of OpenShift.

To fix the above mentioned issue, run the below command and restart the cluster.

# Run as a root user
fgrep -RIl 127.0.0.1:8443 /home/centos/openshift.local.clusterup/ | xargs sed -i 's/127.0.0.1:8443/$HOSTNAME:8443/g'

oc cluster down

oc cluster up --public-hostname=$HOSTNAME/$IP

OpenShift is now set up and would be up and running at https://$HOSTNAME:8443

This repository has sample "Hello World" application written in python using Flask framework. Clone this repository and run the below commands to test and run the python web application locally

#Install all the dependencies
pip install -r requirements.txt

# Run the unit test
python -m unittest test_flask_app

# Run the application locally
python flask_app.py

Access the web page by URL http://localhost:8000.

Before deploying to Openshift, add github credentials as secret and deploy the application using Docker strategy by running the below commands.

# Login into openshift cluster
oc login -u system:admin

# Create secret
oc secret new-basicauth aki-github-cred --username=akilans --prompt

# Link the secrent to Openshift builder
oc secrets link builder aki-github-cred

# Deploy the application usind Docker strategy
oc new-app https://github.com/akilans/openshift-python.git --strategy docker --name python-web-app --source-secret aki-github-cred

# Expose the service. It creates route
oc expose service python-web-app

Create a storage in OpenShift and attach that to python-web-app deployment on /mnt folder using web console. Set "WRITE_FOLDER" as ENV variable with "/mnt" value, so that next version of python application writes files into /mnt folder

Uncomment the below code in flask_app.py file and push it github repo. The below code writes "hello word with timestamp" on file called "hello.txt" under /tmp every 10 seconds interval by default if there is no "WRITE_FOLDER" ENV variable.

# Cron Job at 10 seconds interval

import time

import atexit

import os


from apscheduler.schedulers.background import BackgroundScheduler


def print_hello_timestamp(folder):

    file = open(os.path.join(folder,'hello.txt'), 'a+')

    file.write(time.strftime("%m/%d/%Y %H:%M:%S ")+ "Hello Word \n")

    file.close()


write_folder = str(os.environ.get("WRITE_FOLDER", "/tmp"))

scheduler = BackgroundScheduler()

scheduler.add_job(func=print_hello_timestamp, args=[write_folder], trigger="interval", seconds=10)

scheduler.start()


# Shut down the scheduler when exiting the app

atexit.register(lambda: scheduler.shutdown())

Run the below command to redeploy the new version of python application. This time it will not ask for git credentials as we already binded the git secrets to the build

# Trigger the build
oc start-build python-web-app 

Delete the pod and test whether the data persists or not by logging into container terminal and run the below command.

# Prints content of hello.txt file
tail -f /mnt/hello.txt

We have already installed ansible in our first step and added ansible playbook under ansible folder. Go inside the ansible folder path and run the playbook to test whether the deployed application is up and the corresponding volume mount details. Web application URL [Generated when we exposed the application] is passed as a parameter.

ansible-playbook site.yaml -e web_url=http://python-web-app-myproject.127.0.0.1.nip.io/