Skip to content

Latest commit

 

History

History
70 lines (44 loc) · 2 KB

README.md

File metadata and controls

70 lines (44 loc) · 2 KB
Raw

Description

A fully functioning, flexibly deployable application for visualizations of personal financial data.

Deployment

Local Windows deployment using Chocolate, Docker, and Minikube:

Use a package manager and install docker-machine and kubectl and minikube and run: ./deploy_locally_k8s.ps1

This is a bit deprecated (for now), but you can also run via: docker-compose up --build -d

Local Development:

Install MongoDB, NodeJS and the required NodeJS packages. Then, start a local MongoDB instance bound to port 27017. Finally, run:

node server.js --database_info 127.0.0.1:27017

AWS deployment using Kops:

This is a bit deprecated, due to more interest in GKE as a deployment platform. Tweaking is needed.

Use an AWS Linux machine, clone this repo, and then run: cd Midas/cloud/aws && ./deploy_on_aws_k8s.sh

Google Cloud Platform deployment using Google Kubernetes Engine (GKE):

Ensure billing is enabled for GCP.

Step 1: Spin up K8s cluster that houses the web interface and database:

Use the provided Ansible to spin up the cluster:

cd Midas/cloud/gcp/ansible/playbooks
ansible-playbook -i hosts midas.yaml --tags="midas_gke_cluster"

Connect to the default cluster via Google Cloud shell, then run:

git clone https://github.com/AlexDHoffer/Midas.git
cd Midas/cloud/gcp
./deploy_on_gke.sh

Step 2: Spin up Compute Engine (VM) that gathers data and publishes it to the database through the external service endpoint.

Use the provided Ansible to provision the VM:

cd Midas/cloud/gcp/ansible/playbooks
ansible-playbook -i hosts midas.yaml --tags="midas_data_vm"

External service endpoint portion TBD.

Microsoft Azure deployment:

TBD