How to hybrid Mysql and MongoDB in Play framework?

Which server side framework and database to use for a live price chart website (similar to Bitcoin Wisdom or Bitcoinity)?

• I'm trying to choose a server framework and database for a live price chart website which tracks prices of cryptocoins (similar to Bitcoin), pulling in data from an exchange such as https://btc-e.com/api/documentation. The website will function in a similar manner as http://bitcoinwisdom.com/ or http://bitcoinity.org/. I have several questions: Which web framework (Node.js, Ruby on Rails etc.) will serve such a website well? Does the live ticker count as a true "real time" application for which Node.js will be appropriate? It seems like BTC-e doesn't provide a Websockets based API (unlike, say, Mt.gox, https://en.bitcoin.it/wiki/MtGox/API/Streaming). So will I have to first fetch the data from BTC-e on the server side and then push this data to clients? Or is there any way to fetch this data from BTC-e directly on the client side? Which method is recommended and why? I would like to store all the data fetched from exchanges such as BTC-e in a database for long term storage and to create historical price charts. Which database would you recommend for this purpose? MongoDB, MySQL, Postgres etc.? I'm just starting out in server-side development and would really appreciate some guidance. :)

• Answer:

  Not a complete answer, but might point in the right direction. Take a look at this Firebase example that seems to do the beginning of what you're looking for. http://btcquote.com/ Blog post about it here: https://www.firebase.com/blog/2013-10-25-btcquote-bitcoin-price-ticker.html Looks it is written using Firebase as a DB. In fact, it seems to be using an openly accessible Firebase DB that updates itself in real time. https://publicdata-bitcoin.firebaseio.com/ From the "real-time" side, you firebase + angular has lots of available options. If you want to back it with a database, you can use whatever you're comfortable using. Frankly the relative size of this data is really tiny and should fit in memory on even a small machine.