MySQL

At some point of time every DBA must have come across a situation where he has to enable the general log to debug an issue or even to capture all the queries hitting the DB. In such situations what most DBA’s do is inform the stakeholders that he is  enabling the general log and this might cause a drop in performance and in turn the stakeholders ask a question like “What will the percentage of performance drop?  Can you give us a number?” The easy answer is “It is difficult to provide a number”.  I had wanted to address this, but never really made an effort until I saw the below blog post “PERFORMANCE_SCHEMA VS SLOW QUERY LOG”  http://www.mysqlperformanceblog.com/2014/02/11/performance_schema-vs-slow-query-log/ In that particular blog Peter explains why he chose to use slow query log to capture slow queries instead of using the PERFORMANCE SCHEMA. At some point he talks about the overhead of enabling the PERFORMANCE SCHEMA. He used sysbench tool to calculate the overhead and pro