hive優化之去distinct

count(distinct ),在數據量大的狀況下，容易數據傾斜，由於count(distinct)是按group by 字段分組，按distinct字段排序。web 1.單個distinct Select device_name,count(distinct imei) from TableA group by device_name; 使用group by替換:app Select devi

>>阅读原文<<