Shulou(Shulou.com)06/01 Report--

Background:

A database application migrated the database from Alibaba Cloud rds to Beijing AWS and found that all queries would be slow.

Analysis:

Selected several query statements with problems, compared the execution statements, indexes, execution plans, and found that the execution plans, indexes, and database building statements are all the same.

Is there a problem with database version, server parameters, database parameters?

AWS database version is MySQL 5.7, Alibaba Cloud is MySQL 5.6, and AWS rds is higher in CPU and memory configuration than Alibaba Cloud. There is no significant difference between database parameters.

Is IO a problem, create test RDS, do stress tests on Beijing AWS, overseas AWS, and Aliyun RDS respectively, and discover performance aspects: Beijing aws> Overseas aws> Alibaba Cloud aws, ruled out the IO problem, finally show processlist, located that the database has a timed task in the big table aggregation query operation, the table has about 100 G, there are several queries, actually query for more than 2 hours, not yet finished, this operation seriously polluted the Buffer Pool, causing other query data, soon replaced out of memory.

Reference: www.orczhou.com/index.php/2010/05/innodb-plugin-make-buffer-cache-scan-resistant/

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.