This is a sample application that will be used to teach SheCanCodeIT members how to build a web application using Python and Flask.
Install Python 3
Install Pip
Install Virtualenv (Globally)
$ pip install virtualenv
Clone the repo
Setup the development environment
$ cd vendor-management
$ virtualenv venv
$ . venv/bin/activate
Note: This will setup a virtual environment in your machine.
Install application requirements in your virtual environment
$ pip install -r requirements.txt
Note: For this application, we will be using SQLite.
$ flask db upgrade
If you made changes to the models code, you will need to run the migration script
$ flask db migrate -m 'add comment here'
$ flask db upgrade
$ flask run
Bring the Auth REST API
Another way of interacting with this application is through a REST APIs. To see the information about the REST API, you can access the Swagger docs
- Auth API (http://localhost:5000/u/v1/)
- Product API (http://localhost:5000/p/v1)
- Notes API (http://localhost:5000/n/v1)
- Contacts API (http://localhost:5000/c/v1)
- Attachment API (http://localhost:5000/a/v1)
Run unittest
$ export FLASK_ENV=test
$ flask test
Run test coverage and show coverage reports
$ export FLASK_ENV=test
$ coverage run vms_test_suite.py
$ coverage report --omit="app/*/test/*.py" app/*/*/*.py
$ coverage html --omit="app/*/test/*.py" app/*/*/*.py
The html coverage report will be generated under htmlcov folder. Under that folder launch index.html in your favorite browser.
Before you can do this, you will ned to install docker. You can find instructions on how to install Docker here.
You will also need to create a vms.env file that contains settings for your database.
$ touch vms.env
Add the following settings in the vms.env file.
SERVICE_NAME=Vendor Managetment Service
CURRENT_VERSION=1.0-uat
SECRET_KEY=secretkey
FLASK_ENV=production
POSTGRES_DB=vms
POSTGRES_USER=vms_user
POSTGRES_PASSWORD=vms_user
POSTGRES_HOST=vms-db
POSTGRES_PORT=5432
Build and run the docker containers. docker-compose build will build your application into a docker image. docker-compose up will run your application pointed to a Postgress database.
$ cd deployment/compose
$ docker-compose build
$ docker-compose up
Access the Auth API via this url - http://localhost/u/v1
minikube is a tool that allows to easily runs a local Kubernetes environment. To install it,
$ brew update
$ brew install kubectl
$ brew install minikube
To run it,
$ minikube config set vm-driver hyperkit
$ minikube start
$ minikube dashboard
If you run into some problems, you can easily delete it.
$ minkube delete
$ minikube delete
$ rm /usr/local/bin/minikube
$ rm -rf ~/.minikube
To deploy the application, first create the Postgres database.
$ cd deployment/minikube
$ kubectl apply -f ./persistent-volume.yml
$ kubectl apply -f ./persistent-volume-claim.yml
$ kubectl apply -f ./secret.yml
$ kubectl create -f ./postgres-deployment.yml
$ kubectl create -f ./postgres-service.yml
$ POD_NAME=$(kubectl get pod -l service=postgres -o jsonpath="{.items[0].metadata.name}")
$ kubectl exec $POD_NAME --stdin --tty -- createdb -U vms_user vms
Before we deploy the application, first build the VMS docker image
$ cd ../..
$ docker build -t ealberto/vms-app:latest .
$ docker push ealberto/vms-app:latest
Deploy the application
$ cd deployment/minikube
$ kubectl create -f ./vms-app-deployment.yml
$ kubectl create -f ./vms-app-service.yml
$ minikube addons enable ingress
$ kubectl apply -f ./minikube-ingress.yml
$ echo "$(minikube ip) erwindev.io" | sudo tee -a /etc/hosts
Create the cluster
gcloud container clusters create vms-cluster --zone us-east1-b --machine-type=n1-standard-1 --max-nodes=3 --min-nodes=1
List the cluster
gcloud container clusters list
Scale down the cluster
gcloud container clusters resize vms-cluster --num-nodes 0 --zone us-east1-b
Get credentials for the cluster
gcloud container clusters get-credentials vms-cluster --zone us-east1-b
Upload the database service credentials
kubectl create secret generic vms-cloudsql-instance-credentials --from-file=sql_credentials.json=/Users/ealberto/mystuff/erwindev-vms-db-b1dca3f9d9a1.json
Create the username and password as secrets
kubectl create secret generic vms-cloudsql-db-credentials --from-literal=username=vms_user --from-literal=password=D3qvzsAJHarhELzvK9
Create deployment
cd deployment/gcp
kubectl create -f vms-app-deployment.yml
List deployed apps
kubectl get deployments
Create service
kubectl create -f vms-app-service.yml
List Services
kubectl get services
Create a static IP
gcloud compute addresses create vms-static-ip --global
List IP addresses
gcloud compute addresses list
If working with multiple, context
kubectl config get-contexts
kubectl config use-context [CONTEXT]
For load testing, we will use Locust. The load testing script is located under the loadtest folder. Curerntly, we are only load testing the /u/v1/auth/login api. To run the script,
$ cd loadtest
$ locust -f vmsloadtest.py
Bring up the Locust dashboard by going to http://localhost:8089. Enter the following information,
Number of total users to simulate: 100
Hatch rate: 5
Host: http://localhost:5000
Note: Before we can load test, please start the application. The URL will be what you will provide in the "Host" parameter.