var-sandbox

Value at Risk sandbox scaffolding application

This project is a mixed-mode Python Flask, React, and Apache Spark application. As such it is slightly complicated to assemble and run. If you you wish to try it out, the OpenShift deployment option is recommended as the first pass. If you would prefer to dive into development or desire to run the application outside of OpenShift, please see the local deployment instructions.

Basic architecture

Deployment on OpenShift

The openshift directory contains a few templates for deploying the var-sandbox into an OpenShift project. These templates are self explanatory and create all the necessary objects to build and/or deploy the application.

Please note the image and repository locations that are encoded in the templates, you will want to change these URLs to align with your needs.

The file var-sandbox-setup-list.yaml is the easiest to start with, it will create all the necessary objects in your OpenShift project. It is designed with the idea that you will build the application inside of OpenShift.

for a quick start, these steps will get you up and running:

oc create -f var-sandbox-setup-list.yaml
oc start-build var-sandbox
oc new-app --template var-sandbox

these commands will create the objects, start a build of the application, and then launch it when finished. If all goes well you will have an exposed route to your application.

The file var-sandbox.yaml is very similar to the other template but does not include the build configuration. You should use this file if you wish to consume the pregenerated images for the application and do not wish to build the images in your OpenShift project. It can be started with the following commands:

oc create -f var-sandbox.yaml
oc new-app --template var-sandbox

Local deployment

These instructions are meant for developers and advanced users. If you are building for local deployment you will need a copy of Apache Spark and the wikieod.parquet file that is used by the application. The images referenced in the quick start and the Dockerfile contain the necessary parquet data.

Easy install

The easiest way to use this image is to create a container from the Dockerfile and use that to access the application. This can be done most simply with the following:

$ docker build -t var-sandbox .
... <build output> ...
$ docker run --rm -it -p 8080:8080 var-sandbox

Once running, access http://127.0.0.1:8080/ with your browser.

Advanced install

There are 2 main steps to running this project: compile the React components, and run the Flask application.

To help automate these processes we recommend using a Python virtual environment to setup the Flask requirements, and use the Yarn project to install and compile the React components.

Install the Python requirements

$ pip install -r requirements.txt

Install and build the React components

$ yarn install
$ yarn run build

Run the application with Yarn

Before executing the devel run option, you need to provide 2 environment variables, SPARK_SUBMIT_CMD and WIKIEOD_FILE.

$ export SPARK_SUBMIT_CMD=$HOME/opt/spark/bin/spark-submit
$ export WIKIEOD_FILE=$HOME/tmp/wikieod.parquet
$ yarn run devel