cbonoz / devweek2017

DevWeek 2017 Project

NatureFit

Integrated Clover and NetApp backend DeveloperWeek 2017 code. Honorable Mention - NetApp

NatureFit is a new fitness rewards platform designed to boost gym sales and club membership through a gym perk incentive-based credit system that rewards people for making referalls and sticking to their gym routines. Powered by Clover Point of Sale API, and customer tracking - users can earn points for referring signups at the gym and receive gym credit for buying items like protein bars, shakes, etc. at a potentially-unnmanned food stand located within the gym (that has Clover POS set up). NetApp backend allows high resiliency (even in small datacenter configurations) by using a novel instant-backup strategy for replacing stalling instances on the fly.

In many cases, gym snack inventory is priced excessively high and is not leveraged as well as it could be for gym-goers. This can often lead to gyms having to last-minute sell ites at a discount or just letting them expire.

I suppose the closest analogy would be - ** Think Apple Pay meets Five Stars. **

FiveStars: http://hi.fivestars.com/ (building a custom rewards program for the membership vertical).

Clover API (Customer Tracking and Payment Processing)

Clover API is used for both handling transactions at checkouts. For this we primarily use three API routes. One for creating users, and the other two for tracking rewards inventory (at the stores) and handling payments. The customer profile can be tied to the gym's existing DB of users.

Many enterprises are still cautious about moving their entire data to a single backend provider. Hidden on the backend of this app is a fully-functional (prototype) 2-instance distributed cluster, which handles requests coming in. 2 instances, impressive.

NetApp Mirroring

Utilizing NetApp's Efficient Data Cloning Strategy.

Disaster recovery. Using volume SnapMirror, data is mirrored in real time to another NetApp storage system at a DR facility or secondary data center. If a DR version needs to be made operational, applications are switched over to a newly mirrored OnTap Cloud instance and the traffic is redirected for as long as necessary. When the production site is back online, SnapMirror can transfer the data efficiently back to the production storage systems, and SnapMirror transfers can resume.

We integrated the snap-mirror technology natively into the consumer-side of the app - working in tandem to listen to any latency increases or time outs that occur for an end user. In the case of any anomoly on an instance, we use the agility of NetApp's spin-mirror to mirror the troubled instance and redirect traffic (automatically). This minimizes the notice/impact of end users while minizing the cost for the provider.

Set up instruction:

1. You will need your AWS account set up with instances supported by NetApp's OnTap Cloud technology. This is well documented on the AWS website, so we won't focus on that here.
2. Both server's should be running the NatureFit backend server, with a shared volume.
3. Run MongoDB. The application uses a DB backed by PyMongo technology.

Using NetApp for auto-scaling/DB management. Using Clover for customer management and tracking.

Demo Presentation:

1. Story of NatureFit
2. Failure and instant snapmirror. We have back up instances configured, but not brought online. AWS has autoscaling and step-based scaling but easy to be too conservative or too late (leading to a poor performance period for users while new instances are generated). We take advance of OnTap's quick mirroring to build an instance on demand, minimizing the spin up time and reducing the overall costs by allowing you to be more precise with your scaling triggers.
3. Resync after production back online.

Developers

Backend: Chris Buonocore

FrontEnd: Michael Tran, Andrew Balberchak

DevNotes

Angular CLI: https://github.com/angular/angular-cli

Connecting to Server:

Instance/IP 1: ssh -i serverkey.pem [email protected]

Instance/IP 2: ssh -i serverkey.pem [email protected]

Synchronize local directory with server file: rsync -avz -e "ssh -i serverkey.pem" -a . [email protected]:~/www/naturefit

Running the server: python server.py