MySQL Error_code: 1756

昨天线上发现一个数据库诡异问题，在这里记录下问题的发现以及处理的过程。

研发的同事反馈一个数据库的数据从19：44就没有更新了（该从库线上服务不适用，只是线下查询以及统计数据使用）。但是登录到从库查看发现主从同步正常，负责主从同步的两个进程为yes，主从延迟为0，Slave_SQL_Running_State为Waiting for Slave Workers to free pending events

该状态是什么意思呢？查看MySQL官网解释如下：https://dev.mysql.com/doc/refman/5.7/en/slave-sql-thread-states.html

Waiting for Slave Workers to free pending events

This waiting action occurs when the total size of events being processed by Workers exceeds the size of theslave_pending_jobs_size_max system variable. The Coordinator resumes scheduling when the size drops below this limit. This state occurs only when slave_parallel_workers is set greater than 0.

翻译如下：

这个等待状态会在什么时候发生呢？这个状态只会在slave_parallel_workers设置不为0时，当Workers处理的事件总大小超过了系统参数slave_pending_jobs_size_max设置的值时。当大小低于这个值时，调度器才会恢复调度。

查看错误日志信息

2019-09-02T20:45:40.060392+08:00 6 [ERROR] Slave SQL for channel '': ... The slave coordinator and worker threads are stopped, possibly leaving data in inconsistent state. A restart should restore consistency automatically, although using non-transactional storage for data or info tables or DDL queries could lead to problems. In such cases you have to examine your data (see documentation for details). Error_code: 1756

show variables like '%slave_pending_jobs_size_max%'; 默认为16M