slave_pending_jobs_size_max

自己的复制环境（5.7.10）还没有使用并发复制，所以想尝试一下 :

原参数如下：

| slave_parallel_type          | DATABASE                      |
| slave_parallel_workers       | 0                             |

将参数更改如下：

| slave_parallel_type          | LOGICAL_CLOCK                 |
| slave_parallel_workers       | 4                            |

两天后，备库遇到了一个错误：

[root@localhost data]# cat  localhost.localdomain.err | grep 'Cannot schedule' | more
2018-06-23T01:52:06.212083Z 1284585 [ERROR] Slave SQL for channel 'semir_33': Cannot schedule event Query, relay-log name ./localhost-relay-bin-semir_33.0337
54, position 969425791 to Worker thread because its size 19800765 exceeds 16777216 of slave_pending_jobs_size_max. Error_code: 1864
2018-06-23T01:52:06.212142Z 1284585 [Warning] Slave: Cannot schedule event Query, relay-log name ./localhost-relay-bin-semir_33.033754, position 969425791 to
 Worker thread because its size 19800765 exceeds 16777216 of slave_pending_jobs_size_max. Error_code: 1864

好像是说数据超过了 slave_pending_jobs_size_max 的最大值，查看从库参数当前值，确实是16777216 ，改成 29999104 ，备库能够正常起来了 。

root@localhost :(none)12:00:11>show variables like 'slave_pending_jobs_size_max';
+-----------------------------+----------+
| Variable_name               | Value    |
+-----------------------------+----------+
| slave_pending_jobs_size_max | 16777216 |
+-----------------------------+----------+

 

但是还发现一个问题，有很多note 的信息，并且复制的速度并没有什么提升，暂时改回去了

2018-06-26T02:55:26.836074Z 1285873 [Note] Multi-threaded slave statistics for channel 'semir_33': seconds elapsed = 120; events assigned = 138388481; worker queues filled over overrun level = 14564; waited due a Worker queue full = 0; waited due the total size = 39834; waited at clock conflicts = 94796032330100 waited (count) when Workers occupied = 206047 waited when Workers occupied = 6291101300
2018-06-26T02:59:25.314146Z 1285873 [Note] Multi-threaded slave statistics for channel 'semir_33': seconds elapsed = 239; events assigned = 138404865; worker queues filled over overrun level = 14564; waited due a Worker queue full = 0; waited due the total size = 39834; waited at clock conflicts = 95034306234300 waited (count) when Workers occupied = 206047 waited when Workers occupied = 6291101300
2018-06-26T03:01:43.792496Z 1285873 [Note] Multi-threaded slave statistics for channel 'semir_33': seconds elapsed = 138; events assigned = 138508289; worker queues filled over overrun level = 14564; waited due a Worker queue full = 0; waited due the total size = 39834; waited at clock conflicts = 95171245248000 waited (count) when Workers occupied = 206047 waited when Workers occupied = 6291101300