Hello,
The application I am currently working on is seeing some strange problems seemingly around Connector/J version 5.1.12 and some nasty CPU spikes. We've added some logging statements around com.mysql.jdbc.PreparedStatement.ParseInfo() because we noticed that a large number of our threads were spending a lot of time inside this constructor during some of our spikes. SQL Queries that usually take milliseconds to be parsed by this constructor are sometimes taking upwards of 90 seconds to complete.
Our logging shows us that this constructor sometimes parses a few characters, then stops for 100 to 150 milliseconds, then continues parsing the remaining charaters until it stops again a short time later. This continues until some of larger SQL queries end up taking 10s of seconds to be parsed.
I've searched the MySQL bug tracker for any intermittent problems around this code, but I did not find any.
I will attach the relevant data we've gathered from this logging shortly.
The application I am currently working on is seeing some strange problems seemingly around Connector/J version 5.1.12 and some nasty CPU spikes. We've added some logging statements around com.mysql.jdbc.PreparedStatement.ParseInfo() because we noticed that a large number of our threads were spending a lot of time inside this constructor during some of our spikes. SQL Queries that usually take milliseconds to be parsed by this constructor are sometimes taking upwards of 90 seconds to complete.
Our logging shows us that this constructor sometimes parses a few characters, then stops for 100 to 150 milliseconds, then continues parsing the remaining charaters until it stops again a short time later. This continues until some of larger SQL queries end up taking 10s of seconds to be parsed.
I've searched the MySQL bug tracker for any intermittent problems around this code, but I did not find any.
I will attach the relevant data we've gathered from this logging shortly.