我在HBase中存储数据,它有350万条记录。我正在使用Apache Drill Java应用程序来查询此hbase数据,但简单的查询需要3分钟才能完成。
我分配了drill_direct_memory = 10GB和drill_heap = 8GB。如何获得快速查询性能?我是否遗漏了为获得更好的性能而必须做的任何配置或任何更改?
这是我的示例查询:
SELECT CONVERT_FROM(row_key,' UTF8')AS row_key,CONVERT_FROM(social_data_nlp.details.user_name,' UTF8')AS user_name, TO_CHAR(CAST(social_data_nlp.details.created_time AS TIMESTAMP),' yyyy:MM:dd HH:mm:ss')AS created_time FROM hbase.social_data_nlp WHERE CAST(social_data_nlp.details.created_time AS date)> DATE_SUB(CURRENT_DATE,1)AND(social_data_nlp.details.searched_keyword =' fire' or social_data_nlp.details.searched_keyword =' accident' OR social_data_nlp.details.searched_keyword =' karnataka&# 39; OR social_data_nlp.details.user_name =' aajtak')AND CAST(social_data_nlp.details.keyword_search as boolean)= true ORDER BY social_data_nlp.details.created_time DESC LIMIT 100 OFFSET 0;