CerebralCortex with MySQL Pooling and fixed data replay bug

1 - Added functionality to use MySQL pooling (please set 1 connection as pool-size in configs)
2 - Fixed bug to correctly add datapoints timestamps

Bug fixes

1 - Minio methods were rewritten to make it backward compatible with mCerebrum
2 - Added test-cases
3 - Added some helper methods to get user metadata
4 - Changed json protocol to line protocol for influxdb

Updated Architecture For Scaling issues

Restructured code to avoid nested RDD errors
Changed the way data is stored in Cassandra
Added logging capability
Added Test-cases