标签：
实验步骤
同一台电脑装了两个mysql，端口号不同，mysql5.6安装好之后只有一个my-default的文件，将其中添加一些错误内容，改mysql依然可以启动成功
，说明使用的不是该文件，将my-default.ini改名为my.ini，再向其中添加一些错误内容，则启动失败，则说明，mysql5.6启动时读取的配置文件
名称默认为my.ini
使用框架springmvc+spring+hibernate，但是插入数据时使用的JdbcDaoSupport中的jdbcTemplete，所以在这里hibernate也算是没有使用到，
由于要获取同一个connection对象，所以在这里封装了个方法
public static final ThreadLocal<Connection> cs = new ThreadLocal<Connection>(); public static Connection getConnection(DataSource dataSource) throws SQLException{ Connection c = cs.get(); if(null==c){ c=dataSource.getConnection(); cs.set(c); } return c; }
dao插入数据代码
package com.h3c.itac.alarm.dao; import java.sql.Connection; import java.sql.PreparedStatement; import java.sql.ResultSet; import java.sql.SQLException; import javax.annotation.PostConstruct; import javax.annotation.Resource; import javax.sql.DataSource; import org.springframework.jdbc.core.support.JdbcDaoSupport; import org.springframework.stereotype.Repository; import com.h3c.itac.alarm.po.SyslogAlarm; import com.h3c.itac.util.ConnectionUtil; @Repository public class SyslogAlarmDao extends JdbcDaoSupport{ @Resource public void setDatasource(DataSource dataSource){ this.setDataSource(dataSource); } private long count = 0L; String sql1=""; String sql2=""; PreparedStatement ps1 = null; PreparedStatement ps2 = null; Connection con = null; @PostConstruct public void set() throws SQLException{ sql1 = "insert into alarm(id,adesk_alarm_id,serial_no,level,source,type,device_id,create_time,upload_time,order_id,title,location,customer_id)" +"values(?,?,?,?,?,?,?,?,?,?,?,?,?)"; sql2 = "insert into syslogalarm values(?,?,?,?,?,?,?,?)"; con = ConnectionUtil.getConnection(this.getDataSource()); // con.setAutoCommit(false); ps1 = con.prepareStatement(sql1); ps2 = con.prepareStatement(sql2); PreparedStatement ps3 = con.prepareStatement("select max(id) from alarm"); ResultSet rs = ps3.executeQuery(); while(rs.next()){ count = rs.getLong(1); } System.out.println("alarm id最大值为 ："+count); } public void insert(SyslogAlarm sAlarm) throws SQLException{ count++; ps1.setLong(1,count); ps1.setLong(2,sAlarm.getAlarm().getAdeskAlarmId()); ps1.setString(3,sAlarm.getAlarm().getSerialNo()); ps1.setInt(4, sAlarm.getAlarm().getLevel()); ps1.setInt(5, sAlarm.getAlarm().getSource()); ps1.setInt(6,sAlarm.getAlarm().getType()); ps1.setLong(7, sAlarm.getAlarm().getDevice().getId()); ps1.setLong(8, sAlarm.getAlarm().getCreateTime()); ps1.setLong(9, sAlarm.getAlarm().getUploadTime()); ps1.setLong(10,count); ps1.setString(11, sAlarm.getAlarm().getTitle()); ps1.setString(12, sAlarm.getAlarm().getLocation()); ps1.setLong(13,sAlarm.getAlarm().getDevice().getCustomer().getId()); ps2.setLong(1,count); ps2.setString(2, sAlarm.getDescription()); ps2.setString(4,sAlarm.getReason()); ps2.setString(5,sAlarm.getAdvise()); ps2.setString(3, sAlarm.getVariables()); ps2.setString(6, sAlarm.getLogContent()); ps2.setLong(7,sAlarm.getPriginalLogId()); ps2.setLong(8, count); ps1.addBatch(); ps2.addBatch(); if(count%300==0){ excuteRemainderBatchSQL(); } } public void excuteRemainderBatchSQL() throws SQLException{ ps1.executeBatch(); ps1.clearBatch(); ps2.executeBatch(); ps2.clearBatch(); // con.commit(); // con.setAutoCommit(true); } }
测试类
package com.h3c.itac.alarm.dao; import java.sql.SQLException; import java.util.Date; import org.junit.Test; import org.springframework.context.ApplicationContext; import org.springframework.context.support.ClassPathXmlApplicationContext; import com.h3c.itac.alarm.po.Alarm; import com.h3c.itac.alarm.po.SyslogAlarm; import com.h3c.itac.customer.Customer; import com.h3c.itac.device.Device; public class TestSyslogAlarmDao { public SyslogAlarmDao getSyslogAlarmDao() { ApplicationContext ac = new ClassPathXmlApplicationContext( "applicationContext.xml"); return (SyslogAlarmDao) ac.getBean("syslogAlarmDao"); } @Test public void init(){ ApplicationContext ac = new ClassPathXmlApplicationContext( "applicationContext.xml"); } /**每次清库1800 * 2,自动提交（2.388,2.388,2.871,2.809,2.683）平均2.6278 * 1，手动提交(2.825,2.717,2.294,2.076,2.169)平均2.4162 * 不清库,每次1800 * 2,自动提交（2.981,3.261,3.433,1.888,1.935） * 1，手动提交（2.091，2.293，3.354，1.95，2.371，） * 5000 * 2，自动提交（6.397,5.43,6.584,6.35,4.853） 平均5.9228 * 1,手动提交(4.166，5.913，4.634，2.762，4.962) 平均4.4874 */ @Test public void testInsert() throws SQLException, ClassNotFoundException{ double start = System.currentTimeMillis(); SyslogAlarmDao sad = this.getSyslogAlarmDao(); for(int i=1;i<5001;i++){ Customer c = new Customer(); c.setId(1L); Device d = new Device(); d.setId(1L); d.setCustomer(c); Alarm alarm = new Alarm(); alarm.setAdeskAlarmId((long) i); alarm.setCreateTime(new Date().getTime()); alarm.setDevice(d); alarm.setLevel(1); alarm.setLocation("location "+i); alarm.setSerialNo("serialno "+i); alarm.setSource(1); alarm.setTitle("title "+i); alarm.setType(0); alarm.setUploadTime(new Date().getTime()); SyslogAlarm sa = new SyslogAlarm(); sa.setAdvise("advise "+i); sa.setAlarm(alarm); sa.setDescription("description "+i); sa.setLogContent("logcontent "+i); sa.setPriginalLogId(1); sa.setReason("reason "+i); sa.setVariables("variable "+1); sad.insert(sa); } sad.excuteRemainderBatchSQL(); double end300 = System.currentTimeMillis(); System.out.println((end300-start)/1000); } }
如果只是将my-default.ini文件改名为my.ini文件，而不该内部配置，则在执行testInsert的时候速度超慢，即便是插入600条也需要将近二十多秒
然后找来其他版本的my.ini文件，替换掉改名后的my.ini文件，则速度比较快,可以想到问题就出在之前版本的my.ini和改名后的my.ini的文件配置内容上
然后用Compare对比改名前后的my.ini文件内容使用Compare辅助工具进行测试，（使用Compare对比在改名的my.ini文件中不存在的内容，将之前版本的my.ini文件内容逐行添加到改名后文件中，停止mysql服务，保存修改后文件，清空数据库，运行testInsert方法，观察运行时间，如果这条没有效果，继续添加下条配置）
后来就发现有一个配置起了作用
# If set to 1, InnoDB will flush (fsync) the transaction logs to the
# disk at each commit, which offers full ACID behavior. If you are
# willing to compromise this safety, and you are running small
# transactions, you may set this to 0 or 2 to reduce disk I/O to the
# logs. Value 0 means that the log is only written to the log file and
# the log file flushed to disk approximately once per second. Value 2
# means the log is written to the log file at each commit, but the log
# file is only flushed to disk approximately once per second.
innodb_flush_log_at_trx_commit = 2
一、当该配置为2的时候，代码还是如上，自动提交事务，则插入数据速度明显提升，和使用之前版本的my.ini文件几乎是相同的，
二、当该配置为1的时候，代码依然如上，自动提交事务，则插入数据速度又变得明显超慢
三、当该配置依然为1的时候，自动提交事务，即将上述代码con.setAutoCommit(false);和con.commit();的注释去掉，则插入数据也有明显提升，和第一种情况差不多，当插入数据过多时比第一种要快多一些
四、当该配置为2的时候，自动提交事务，即将上述代码con.setAutoCommit(false);和con.commit();的注释去掉，插入速度也会有明显提升。
五、去掉innodb_flush_log_at_trx_commit = 2配置，设置为手动提交事务，插入速度也会明显提升
六、去掉innodb_flush_log_at_trx_commit = 2配置，自动提交事务，速度又是明显超慢。
综上，如果用的是mysql5.6版本，要更改默认配置，目前我们采用的是将my-default.ini文件名改成，my.ini文件名,这样就可以在其中进行文件配置修改
通过上面测试结果也可看出，不管是配置为1或者2，只要是手动提交事务，则都可一提高插入速度，这个是在jdbc中进行的测试，至于hibernate需要就不清楚了
my.ini
# For advice on how to change settings please see # http://dev.mysql.com/doc/refman/5.6/en/server-configuration-defaults.html # *** DO NOT EDIT THIS FILE. It‘s a template which will be copied to the # *** default location during install, and will be replaced if you # *** upgrade to a newer version of MySQL. [mysqld] # generic configuration options port = 3306 socket = /tmp/mysql.sock character-set-server=utf8 innodb_print_all_deadlocks=1 innodb_locks_unsafe_for_binlog=1 expire_logs_days=2 event_scheduler=1 # If set to 1, InnoDB will flush (fsync) the transaction logs to the # disk at each commit, which offers full ACID behavior. If you are # willing to compromise this safety, and you are running small # transactions, you may set this to 0 or 2 to reduce disk I/O to the # logs. Value 0 means that the log is only written to the log file and # the log file flushed to disk approximately once per second. Value 2 # means the log is written to the log file at each commit, but the log # file is only flushed to disk approximately once per second. innodb_flush_log_at_trx_commit = 2
标签：
原文地址：http://www.cnblogs.com/qq931399960/p/4554561.html
