Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Facebook app for architecture discussion
branch: master

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.
create
lonelytest-sdk
search
take
README
comic.ttf
index.php
texture.jpg
uHuman.php

README

I have been into the industry for about an year. Long enough to understand the importance of scalability, but not enough to actually come up with the right solution.

The problem here is simple - A facebook app which lets you run a chosen test, the result (just a random selection from the available options) is posted on user's facebook wall. A number of tests can be created but that won't be frequent. Moreover, tags are associated with a test so that search can be conducted.

Current architecture for the demo is simple
Each test has a corresponding row in LonelyTest table (Id|Question|Options|Tags|Used). Options and Tags are being stored as delimited text.

Assumption:
- Some tests will be popular, so their data must be cached using some solution like memcachedb.
- Tests will be less frequently created than taken.

Constraints: Best available resource is my AWS free tier
Something went wrong with that request. Please try again.