【讨论中】Lost connection to MySQL server during query ( server_errno=2013)

hsu_wp

线上slave机器大量报这个2013错误，每天一条，网上查了下别人说的这种错误：一般因网络问题导致的可能性比较大，另外，反向解析 和max_allowed_packet 不一致也可能会导致这种问题。
Mysql的Error.log 中记录的Error 2013貌似是个很奇怪的错误，它并不影响复制本身，虽然error log中有大量的这类错误记录，我们检查了，发现并没有造成 Master ，Slave 数据不一致的情况。
但这个怎感觉很烦人，也没有找到合适的解决办法，

这个帖子也说道类似的情况，http://bugs.mysql.com/bug.php?id=5588，但是我感觉也没说的很透彻，

有没有兄弟也报类似的错误，或者类似的情况，你们是怎么解决的，怎么理解的？

hsu_wp

附部分错误：
130204  5:00:02 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130204  5:00:02 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000099', position 420402
130204  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000099' at position 420402
130205  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130205  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000102', position 2133703
130205  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000102' at position 2133703
130206  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130206  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000102', position 3077466
130206  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000102' at position 3077466
130207  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130207  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000102', position 4313793
130207  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000102' at position 4313793
130208  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130208  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000102', position 5715859
130208  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000102' at position 5715859
130209  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130209  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000102', position 7117936
130209  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000102' at position 7117936
130210  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130210  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000102', position 8520313
130210  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000102' at position 8520313
130211  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130211  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000102', position 9922373
130211  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000102' at position 9922373
130212  5:00:01 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
130212  5:00:01 [Note] Slave I/O thread exiting, read up to log 'chcp-cnc-db01-bin.000103', position 838611
130212  5:00:03 [Note] Slave I/O thread: connected to master 'cy_repl@10.182.0.14:3306',replication started in log 'chcp-cnc-db01-bin.000103' at position 838611

hsu_wp

这篇帖子也说到类似的问题:
http://www.itpub.net/thread-1506684-1-1.html

lengkuqiufeng

我最近也遇到相同的错误，不知道楼主是如何解决的呢？


   

chinafenghao

@hsu_wp
楼主,你发现没有。
1、你每天出错的时间几乎一样。这里确定是某个固定因素触发的,比如计划任务。
2、这个错误一般都是网络问题引发的。

所以你可以考虑排查这两台机器每天5点的固定操作,5点时候数据库服务器的各项性能指标等，找出什么原因导致网络异常，最后影响到slave的报错。,

doondroar

我和master-master的也有类似的问题，但是replication又是好的，贴一下错误，这里：
master:3306
140808 14:08:06 [ERROR] Slave I/O: error reconnecting to master 'repli@slave:3306' - retry-time: 60  retries: 86400, Error_code: 2003
140808 14:20:31 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
slave:3306
140808 14:08:06 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
140808 14:20:31 [ERROR] Slave I/O: error reconnecting to master 'repli@master:3306' - retry-time: 60  retries: 86400, Error_code: 2003
话说是个bug

saup007

我使用innobackupex备份slave时，今天遇到过一次。。