Using SnappyData as the database store

[fire]

Snappydata is:

1. spark which allows us to use machine learning algorithms from spark
2. a sql database which allows us to use standard sql
3. apache licensed which is the same license as octo.ai
4. approximate queries allow us to run sampling on the source data

[dipankar]

Hey @fire Thanks for putting up the issue!

[dipankar]

We can finally consider taking this up! Adding more dev firepower.

We just wrapped up Mongo, HBase support as that was in demand.

[fire]

Great! Can't wait to view your progress.

[fire]

If this helps I found:

• https://github.com/chyh1990/jruby-spark
• https://github.com/ondra-m/ruby-spark
• https://github.com/jruby/activerecord-jdbc-adapter

[dipankar]

Thanks, that is the direction we are taking on this!

[fire]

How has it been going? I also had trouble finding a ruby library for interfacing with Spark and SnappyData.

[dipankar]

Yes, however we are looking at alternate mechanisms to integrate.

Maybe build more specific data models just for snappy.

Though that will make quite a bit of the other code unusable (dashboard, etc.)