4 ms·
Generally good idea, analytical queries will be faster than on rdbms. You problem in strange tools and format. Hive = map reduce, JSON = scan for all data. You
by Triffids 8y ago
Generally good idea, analytical queries will be faster than on rdbms. You problem in strange tools and format. Hive = map reduce, JSON = scan for all data. You need to replace map-reduce (Hive) to Impala or Spark 2.x and JSON to columnar format: Parquet or ORC.