standardize the migration and DB setup process

[OriHoch]

Currently there are duplicated migrations:

• Knex - https:/Midburn/Spark/tree/master/migrations
• .sql files - https:/Midburn/Spark/tree/master/sql

Also, there is no standard simple way to run migration (at least from what I could see)

TODO

• decide on which migration tool we are using and remove the other one
• create a standardized way to run the migrations - e.g. a bash file bin/run_mgrations.sh

[ripper234]

Odd. I translated the major schema to knex (because it's compatible with other databases including SQLite I introduced for testing). I thought I deleted the main schema.sql file, but apparently not.

Need to:

1. Synchronize any changes from schema.sql into migrations/20161208025453_init.js (knex)
2. Delete schema.sql
3. Check with the team if anyone needs the other SQL files in the sql/ folder (basically they shouldn't ... the initial data should all be in one file right now, and that file's format should be knex).
4. Mirgate the other files into 20161208025453_init.js knex + delete
5. Document how to work with knex migrations and inform the teams & team leads.

[ripper234]

BTW regarding your second TODO above: knex is the standard, pretty well documented.
If there's a need for a mini-script (not sure that there is), we can add it to the scripts section of package.json ... but I don't think we need an external bash script.

[OriHoch]

Cool, I'll do it

[OriHoch]

@ripper234 - opened a PR which switches to Knex, please review - #15

[OriHoch]

Follow-up tasks:

• update documentation about how to use Knex and inform the team - update documentation about how to use Knex and inform the team #16
• use Sqlite by default for local development - should use sqlite for local development #17