mysql8性能参数查看及优化

程序猿（攻城狮）2024-10-27 11:50

mysql的整体性能和基本环境和参数配置都有很大的关系，在特定项目环境下，mysql的运行环境就是确定的，比如cpu配置，内存大小，磁盘类型等。为了充分利用环境提供的处理能力，就有必要对相关参数进行优化配置，以达到性能优化的目的。

当然，参数值的优化配置也是需要和具体的环境信息进行综合考虑的，比如对于16G内存的服务器和8G以及4G内存的服务器，优化后的参数值通常是不一样的。除此之外，还需要根据业务特定进行优化，比如对于需要快速响应的业务需求和批处理业务需求，可能在参数优化上的侧重点就会不一样。

下面对相关参数，分别列举了云生产环境、测试环境和本地开发环境，对比之后，就可以找到合适的参数配置：

1. innodb_buffer_pool_size（*）

innodb_buffer_pool_size一般为内存的70%～80%；

innodb_buffer_pool_size = 256MB

查看方法：

复制代码 
-- 云生产环境 cloud mysql:
mysql> SELECT 'innodb_buffer_pool_size' AS name, @@innodb_buffer_pool_size/1024/1024 AS 'value(MB)';
+-------------------------+---------------+
| name                    | value(MB)     |
+-------------------------+---------------+
| innodb_buffer_pool_size | 2868.00000000 |
+-------------------------+---------------+
1 row in set (0.00 sec)

-- 测试环境 192.168.4.125:
mysql> SELECT 'innodb_buffer_pool_size' AS name, @@innodb_buffer_pool_size/1024/1024 AS 'value(MB)';
+-------------------------+--------------+
| name                    | value(MB)    |
+-------------------------+--------------+
| innodb_buffer_pool_size | 128.00000000 |
+-------------------------+--------------+
1 row in set (0.02 sec)

-- 本地开发环境 localhost
mysql> SELECT 'innodb_buffer_pool_size' AS name, @@innodb_buffer_pool_size/1024/1024 AS 'value(MB)';
+-------------------------+------------+
| name                    | value(MB)  |
+-------------------------+------------+
| innodb_buffer_pool_size | 8.00000000 |
+-------------------------+------------+
1 row in set

2. join_buffer_size

join_buffer_size = 16M

复制代码 
-- 云生产环境 cloud mysql:

mysql> SELECT 'join_buffer_size' AS name, @@join_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| join_buffer_size | 0.34375000 |
+------------------+------------+
1 row in set (0.00 sec)

-- 测试环境 192.168.4.125:

mysql> SELECT 'join_buffer_size' AS name, @@join_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| join_buffer_size | 0.25000000 |
+------------------+------------+
1 row in set (0.00 sec)

3. query_cache_size

query_cache_size = 64M

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'query_cache_size' AS name, @@query_cache_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| query_cache_size | 3.00000000 |
+------------------+------------+
1 row in set (0.00 sec)


-- 测试环境192.168.4.125:

mysql> SELECT 'query_cache_size' AS name, @@query_cache_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| query_cache_size | 1.00000000 |
+------------------+------------+
1 row in set, 1 warning (0.00 sec)

4. sort_buffer_size

配置文件：

sort_buffer_size = 16M

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'sort_buffer_size' AS name, @@sort_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| sort_buffer_size | 0.68750000 |
+------------------+------------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'sort_buffer_size' AS name, @@sort_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| sort_buffer_size | 0.50000000 |
+------------------+------------+
1 row in set (0.00 sec)

-- 本地开发环境localhost

mysql> SELECT 'sort_buffer_size' AS name, @@sort_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| sort_buffer_size | 0.25000000 |
+------------------+------------+
1 row in set

5. read_buffer_size

配置文件：

read_buffer_size = 16M

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'read_buffer_size' AS name, @@read_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| read_buffer_size | 0.68750000 |
+------------------+------------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'read_buffer_size' AS name, @@read_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| read_buffer_size | 0.25000000 |
+------------------+------------+
1 row in set (0.00 sec)

-- 本地开发环境localhost

mysql> SELECT 'read_buffer_size' AS name, @@read_buffer_size/1024/1024 AS 'value(MB)';
+------------------+------------+
| name             | value(MB)  |
+------------------+------------+
| read_buffer_size | 0.06250000 |
+------------------+------------+
1 row in set

6. max_connections（*）

配置文件：

max_connections = 768

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'max_connections' AS name, @@max_connections AS 'value';
+-----------------+-------+
| name            | value |
+-----------------+-------+
| max_connections |  1712 |
+-----------------+-------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'max_connections' AS name, @@max_connections AS 'value';
+-----------------+-------+
| name            | value |
+-----------------+-------+
| max_connections |   151 |
+-----------------+-------+
1 row in set (0.00 sec)

-- 本地开发环境localhost

mysql> SELECT 'max_connections' AS name, @@max_connections AS 'value';
+-----------------+-------+
| name            | value |
+-----------------+-------+
| max_connections |   151 |
+-----------------+-------+
1 row in set

7. max_allowed_packet（*）

max_allowed_packet = 4M

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'max_allowed_packet' AS name, @@max_allowed_packet/1024/1024 AS 'value(MB)';
+--------------------+---------------+
| name               | value(MB)     |
+--------------------+---------------+
| max_allowed_packet | 1024.00000000 |
+--------------------+---------------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'max_allowed_packet' AS name, @@max_allowed_packet/1024/1024 AS 'value(MB)';
+--------------------+------------+
| name               | value(MB)  |
+--------------------+------------+
| max_allowed_packet | 1.00000000 |
+--------------------+------------+
1 row in set (0.00 sec)

-- 本地开发环境localhost

mysql> SELECT 'max_allowed_packet' AS name, @@max_allowed_packet/1024/1024 AS 'value(MB)';
+--------------------+------------+
| name               | value(MB)  |
+--------------------+------------+
| max_allowed_packet | 4.00000000 |
+--------------------+------------+
1 row in set

8. key_buffer_size

key_buffer_size = 128M

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'key_buffer_size' AS name, @@key_buffer_size/1024/1024 AS 'value(MB)';
+-----------------+-------------+
| name            | value(MB)   |
+-----------------+-------------+
| key_buffer_size | 16.00000000 |
+-----------------+-------------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'key_buffer_size' AS name, @@key_buffer_size/1024/1024 AS 'value(MB)';
+-----------------+-------------+
| name            | value(MB)   |
+-----------------+-------------+
| key_buffer_size | 16.00000000 |
+-----------------+-------------+
1 row in set (0.00 sec)

-- 本地开发环境localhost

mysql> SELECT 'key_buffer_size' AS name, @@key_buffer_size/1024/1024 AS 'value(MB)';
+-----------------+------------+
| name            | value(MB)  |
+-----------------+------------+
| key_buffer_size | 8.00000000 |
+-----------------+------------+
1 row in set

9. back_log（*）

back_log = 500

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'back_log' AS name, @@back_log AS 'value';
+----------+-------+
| name     | value |
+----------+-------+
| back_log |  3000 |
+----------+-------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'back_log' AS name, @@back_log AS 'value';
+----------+-------+
| name     | value |
+----------+-------+
| back_log |    80 |
+----------+-------+
1 row in set (0.00 sec)

-- 本地开发环境localhost

mysql> SELECT 'back_log' AS name, @@back_log AS 'value';
+----------+-------+
| name     | value |
+----------+-------+
| back_log |    80 |
+----------+-------+
1 row in set

10. innodb_flush_log_at_trx_commit

innodb_flush_log_at_trx_commit = 0

事务日志写盘配置

N=0 每隔一秒，把事务日志缓存区的数据写到日志文件中，以及把日志文件的数据刷新到磁盘上；

N=1 每个事务提交时候，把事务日志从缓存区写到日志文件中，并且刷新日志文件的数据到磁盘上；

N=2 每事务提交的时候，把事务日志数据从缓存区写到日志文件中；

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'innodb_flush_log_at_trx_commit' AS name, @@innodb_flush_log_at_trx_commit AS 'value';
+--------------------------------+-------+
| name                           | value |
+--------------------------------+-------+
| innodb_flush_log_at_trx_commit |     1 |
+--------------------------------+-------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'innodb_flush_log_at_trx_commit' AS name, @@innodb_flush_log_at_trx_commit AS 'value';
+--------------------------------+-------+
| name                           | value |
+--------------------------------+-------+
| innodb_flush_log_at_trx_commit |     1 |
+--------------------------------+-------+
1 row in set (0.00 sec)

-- 本地开发环境localhost

mysql> SELECT 'innodb_flush_log_at_trx_commit' AS name, @@innodb_flush_log_at_trx_commit AS 'value';
+--------------------------------+-------+
| name                           | value |
+--------------------------------+-------+
| innodb_flush_log_at_trx_commit |     1 |
+--------------------------------+-------+
1 row in set

11. sync_binlog

sync_binlog = N：

N>0 每向二进制日志文件写入N条SQL或N个事务后，则把二进制日志文件的数据刷新到磁盘上；

N=0 不主动刷新二进制日志文件的数据到磁盘上，而是由操作系统决定；

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'sync_binlog' AS name, @@sync_binlog AS 'value';
+-------------+-------+
| name        | value |
+-------------+-------+
| sync_binlog |     1 |
+-------------+-------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'sync_binlog' AS name, @@sync_binlog AS 'value';
+-------------+-------+
| name        | value |
+-------------+-------+
| sync_binlog |     1 |
+-------------+-------+
1 row in set (0.00 sec)

12. innodb_buffer_pool_instances（*）

mysql innodb的性能与缓冲池的大小有很大的关系。可以通过以下两个参数进行调整。

innodb_buffer_pool_size 和 innodb_buffer_pool_instances

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'innodb_buffer_pool_instances' AS name, @@innodb_buffer_pool_instances AS 'value';
+------------------------------+-------+
| name                         | value |
+------------------------------+-------+
| innodb_buffer_pool_instances |     8 |
+------------------------------+-------+
1 row in set (0.00 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'innodb_buffer_pool_instances' AS name, @@innodb_buffer_pool_instances AS 'value';
+------------------------------+-------+
| name                         | value |
+------------------------------+-------+
| innodb_buffer_pool_instances |     1 |
+------------------------------+-------+
1 row in set (0.00 sec)

13. innodb_log_buffer_size

innodb_log_buffer_size = 2M

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'innodb_log_buffer_size' AS name, @@innodb_log_buffer_size/1024/1024 AS 'value(MB)';
+------------------------+------------+
| name                   | value(MB)  |
+------------------------+------------+
| innodb_log_buffer_size | 8.00000000 |
+------------------------+------------+
1 row in set (0.01 sec)

-- 测试环境192.168.4.125:

mysql> SELECT 'innodb_log_buffer_size' AS name, @@innodb_log_buffer_size/1024/1024 AS 'value(MB)';
+------------------------+-------------+
| name                   | value(MB)   |
+------------------------+-------------+
| innodb_log_buffer_size | 16.00000000 |
+------------------------+-------------+
1 row in set (0.00 sec)

14. tmp_table_size

tmp_table_size = 256M

复制代码 
-- 云生产环境cloud mysql:

mysql> SELECT 'tmp_table_size' AS name, @@tmp_table_size/1024/1024 AS 'value(MB)';
+----------------+------------+
| name           | value(MB)  |
+----------------+------------+
| tmp_table_size | 2.00000000 |
+----------------+------------+
1 row in set (0.01 sec)

-- 测试环境 192.168.4.125:

mysql> SELECT 'tmp_table_size' AS name, @@tmp_table_size/1024/1024 AS 'value(MB)';
+----------------+-------------+
| name           | value(MB)   |
+----------------+-------------+
| tmp_table_size | 16.00000000 |
+----------------+-------------+
1 row in set (0.00 sec)

15. 优化配置

为了能够让配置的参数能够在mysql重启后仍然有效，建议将优化参数配置在mysql的配置文件中，这样，每次mysql重启后，都能使用优化参数进行运行。

这里以8G内存服务器为例配置的优化参数。

复制代码 
innodb_buffer_pool_size = 4096MB
innodb_buffer_pool_instances = 8

max_connections = 1200
max_allowed_packet = 8M

back_log = 1000
相关推荐
老纪的技术唠嗑局1 小时前
超级干货分享：中小企业使用 OceanBase 实践经验汇总
数据库
CLOUD ACE1 小时前
谷歌云代理｜零售商Target 如何利用 Spanner Graph 提升零售发现体验并将数据库维护成本降低 50%
数据库·零售
AI大模型-小雄1 小时前
Codex写分页接口为什么越翻越慢？用Cursor Pagination解决重复与漏数据
大数据·数据库·elasticsearch·搜索引擎·chatgpt·后端开发·codex
少晓年2 小时前
从 MySQL 迁移到人大金仓 KingbaseES：完整指南与实践
数据库·mysql
SelectDB2 小时前
Apache Doris AI RAG 实战：从基础 RAG 到知识图谱增强的技术能力与选型
数据库
SelectDB3 小时前
电商企业 PostgreSQL 迁移 Apache Doris：80TB 分析数据统一平台技术能力与实践
数据库
SelectDB3 小时前
阶跃星辰 Agent 可观测：Apache Doris / SelectDB 的技术能力与实践
数据库
SelectDB3 小时前
ApacheDoris Iceberg V3 湖仓 DML：Apache Doris / SelectDB 的技术能力与实践
数据库
SelectDB3 小时前
ApacheDoris Python UDF：SQL 调用 Python 的技术能力、选型对比与实践
数据库
Crazy________3 小时前
Redis03:持久化存储，大key分析，主从复制及哨兵模式
数据库·redis·容器
热门推荐
01VSCode+CodeX扩展，无法启动02GitHub 镜像站点03国内可直接用、免费额度/永久免费的大模型API清单（含 SiliconFlow、火山、阿里、智谱、百度、Kimi、DeepSeek、DMXAPI 等）042026 国产 AI 大模型横评：DeepSeek、通义千问、Kimi、文心一言、星火、豆包谁更能打？05微信历史版本含下载地址（ Windows PC | 安卓 | MAC ）及设置微信不更新06AI 编程 IDE 全景解析 2026：Agent 全面接管开发链路07历年考研数学一、数学二、数学三真题试卷及答案PDF08如何新建文件夹？ 电脑新建文件夹的4种方法09Agnes AI 免费 API 接入指南：文本、生图、生视频，一套接口全免费102026年8月4日科技热点深度解析：AI大模型群雄逐鹿、卫星互联网组网提速、半导体封装材料革命