hive优化之去distinct

count(distinct ),在数据量大的状况下，容易数据倾斜，由于count(distinct)是按group by 字段分组，按distinct字段排序。web 1.单个distinct Select device_name,count(distinct imei) from TableA group by device_name; 使用group by替换:app Select devi

>>阅读原文<<