在 GreatSQL 支持一种新的优化Hint,名字叫SET_VAR,这个特性支持用户在查询语句里修改 GreatSQL 数据库的一些会话变量,当然修改只是对当前查询会话生效,不会影响到其他会话。
SET_VAR语法
SET_VAR这个hint用于临时设置系统变量的会话值(在单个语句的持续时间内有效)
SET_VAR的用法: SET_VAR(var_name=value)
var_name是被临时修改的会话变量名,value是会话变量的取值
greatsql> SELECT @@unique_checks;SELECT /*+ SET_VAR(unique_checks=OFF) */ @@unique_checks;SELECT @@unique_checks;
SELECT /*+ SET_VAR(sort_buffer_size = 16M) */ name FROM people ORDER BY name;
INSERT /*+ SET_VAR(foreign_key_checks=OFF) */ INTO t2 VALUES(2);
SELECT /*+ SET_VAR(optimizer_switch = 'mrr_cost_based=off') */ 1;
GreatSQL 8.0 之前的操作方法
在 GreatSQL 8.0 之前要对一个查询进行会话变量修改,需要怎么操作:
1.查询之前的系统变量
greatsql> SELECT @@optimizer_switch;
2.备份系统变量
greatsql> SET @old_optimizer_switch = @@optimizer_switch;
3.设置新的变量
greatsql> SET optimizer_switch='index_merge=off';
4.运行查询语句
greatsql> SELECT empno,ename,deptno from emp limit 1;
5.恢复之前的系统变量
greatsql> SET optimizer_switch = @old_optimizer_switch;
是不是有点繁琐,现在我们使用SET_VAR这个新特性,很方便的就可以做这个操作了。
GreatSQL 8.0 的操作方法
greatsql>explain SELECT empno,ename,deptno FROM emp WHERE deptno=10 or ename='CLARK';
+----+-------------+-------+------------+-------------+------------------+------------------+---------+------+------+----------+--------------------------------------------+
| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+----+-------------+-------+------------+-------------+------------------+------------------+---------+------+------+----------+--------------------------------------------+
| 1 | SIMPLE | emp | NULL | index_merge | deptno,idx_ename | deptno,idx_ename | 5,63 | NULL | 4 | 100.00 | Using union(deptno,idx_ename); Using where |
+----+-------------+-------+------------+-------------+------------------+------------------+---------+------+------+----------+--------------------------------------------+
1 row in set, 1 warning (0.01 sec)
从执行计划上看,SQL语句使用了索引合并(type=index_merge),如果不想该sql使用索引合并,则可以通过SET_VAR进行控制。
greatsql>explain SELECT /*+ SET_VAR(optimizer_switch='index_merge=off') */ empno,ename,deptno FROM emp WHERE deptno=10 or ename='CLARK';
+----+-------------+-------+------------+------+------------------+------+---------+------+------+----------+-------------+
| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+----+-------------+-------+------------+------+------------------+------+---------+------+------+----------+-------------+
| 1 | SIMPLE | emp | NULL | ALL | deptno,idx_ename | NULL | NULL | NULL | 14 | 38.10 | Using where |
+----+-------------+-------+------------+------+------------------+------+---------+------+------+----------+-------------+
1 row in set, 1 warning (0.00 sec
这个新特性是不是很方便呢,之前由于优化器的某些设置,少量sql语句选择了错误的执行计划,导致查询语句性能低下,又不能随意更改线上数据库的变量,有了SET_VAR这个新特性,对于这种情况,可以考虑在查询语句中使用set_var优化这条语句。
我们知道,使用hash jion时,会使用到join buffer,join buffer的大小由join_buffer_size控制,其默认值为256k,哈希连接不能使用超过此数量的内存。当哈希连接所需的内存超过可用量时,GreatSQL将使用磁盘上的文件来处理此问题,使用到了磁盘文件,性能会下降,如果只想针对单条语句设置join buffer就可以使用SET_VAR。
CREATE TABLE t1 (c1 INT, c2 INT);
CREATE TABLE t2 (c1 INT, c2 INT);
CREATE TABLE t3 (c1 INT, c2 INT);
分别对t1,t2,t3 插入100万,200万,300万数据
greatsql> SET @@cte_max_recursion_depth = 99999999;
greatsql> INSERT INTO t1
WITH recursive t AS (
SELECT 1 AS c1 ,1 AS c2
UNION ALL
SELECT t.c1+1,t.c1*2
FROM t
WHERE t.c1 <1000000
)
SELECT * FROM t;
Query OK, 1000000 rows affected (10.63 sec)
Records: 1000000 Duplicates: 0 Warnings: 0
greatsql> SELECT @@join_buffer_size;
+--------------------+
| @@join_buffer_size |
+--------------------+
| 262144 |
+--------------------+
1 row in set (0.00 sec)
greatsql> SELECT * FROM t1
JOIN t2 ON (t1.c1 = t2.c1 AND t1.c2 < t2.c2)
JOIN t3 ON (t2.c1 = t3.c1);
Empty set (6.91 sec)
greatsql> SELECT /*+ SET_VAR(join_buffer_size=16777216) */ * FROM t1
JOIN t2 ON (t1.c1 = t2.c1 AND t1.c2 < t2.c2)
JOIN t3 ON (t2.c1 = t3.c1);
Empty set (5.87 sec)
注意事项
1、并非所有会话变量都允许与SET_VAR一起使用。如果设置不支持用SET_VAR更改的系统变量,则会出现警告。
greatsql> SELECT /*+ SET_VAR(collation_server = 'utf8mb4') */ 1;
+---+
| 1 |
+---+
| 1 |
+---+
1 row in set, 1 warning (0.01 sec)
greatsql> SHOW WARNINGS\G
*************************** 1. row ***************************
Level: Warning
Code: 3637
Message: Variable 'collation_server' cannot be set using SET_VAR hint.
1 row in set (0.00 sec)
2、SET_VAR语法只允许设置单个变量,但可以给出多个提示来设置多个变量:
SELECT /*+ SET_VAR(optimizer_switch = 'mrr_cost_based=off')
SET_VAR(max_heap_table_size = 1G) */ 1;
3、如果没有这个系统变量或变量值不正确,则忽略SET_VAR提示并发出警告
SELECT /*+ SET_VAR(max_size = 1G) */ 1;
SELECT /*+ SET_VAR(optimizer_switch = 'mrr_cost_based=yes') */ 1;
第1条语句没有 max_size 这个变量,语句2 的mrr_cost_based= on或者off, 企图将其设置为 yes是错误的,这两个语句的 hint 都会被忽略,并产生一个warning。
greatsql> SELECT /*+ SET_VAR(max_size = 1G) */ 1;
+---+
| 1 |
+---+
| 1 |
+---+
1 row in set, 1 warning (0.00 sec)
greatsql> show warnings;
+---------+------+---------------------------------------------+
| Level | Code | Message |
+---------+------+---------------------------------------------+
| Warning | 3128 | Unresolved name 'max_size' for SET_VAR hint |
+---------+------+---------------------------------------------+
1 row in set (0.00 sec)
greatsql> SELECT /*+ SET_VAR(optimizer_switch = 'mrr_cost_based=yes') */ 1;
+---+
| 1 |
+---+
| 1 |
+---+
1 row in set, 1 warning (0.00 sec)
greatsql> show warnings;
+---------+------+-------------------------------------------------------------------------------+
| Level | Code | Message |
+---------+------+-------------------------------------------------------------------------------+
| Warning | 1231 | Variable 'optimizer_switch' can't be set to the value of 'mrr_cost_based=yes' |
+---------+------+-------------------------------------------------------------------------------+
1 row in set (0.00 sec)
4、SET_VAR提示只允许在语句级别使用。如果在子查询中使用,则会被忽略并给出警告。
复制会忽略复制语句中的SET_VAR,以避免潜在的安全问题。
SET_VAR支持的变量
SET_VAR只是对部分变量可以用的,整理了GreatSQL主要支持的变量供参考:
- bulk_insert_buffer_size
- default_table_encryption
- default_tmp_storage_engine
- div_precision_increment
- end_markers_in_json
- eq_range_index_dive_limit
- foreign_key_checks
- group_concat_max_len
- internal_tmp_mem_storage_engine
- join_buffer_size
- lock_wait_timeout
- max_error_count
- max_execution_time
- max_heap_table_size
- max_join_size
- max_length_for_sort_data
- max_points_in_geometry
- max_seeks_for_key
- max_sort_length
- optimizer_prune_level
- optimizer_search_depth
- optimizer_switch
- optimizer_trace_max_mem_size
- range_alloc_block_size
- read_buffer_size
- read_rnd_buffer_size
- secondary_engine_cost_threshold
- select_into_buffer_size
- select_into_disk_sync
- select_into_disk_sync_delay
- show_create_table_skip_secondary_engine
- sort_buffer_size
- sql_auto_is_null
- sql_big_selects
- sql_buffer_result
- sql_mode
- sql_require_primary_key
- sql_safe_updates
- sql_select_limit
- time_zone (≥ 8.0.17)
- timestamp
- tmp_table_size
- unique_checks
- updatable_views_with_limit
- use_secondary_engine
- windowing_use_high_precision
参考文档
- https://dev.mysql.com/doc/refman/8.0/en/optimizer-hints.html#optimizer-hints-set-var
- https://dev.mysql.com/doc/refman/8.0/en/server-system-variables.html
- https://dev.mysql.com/doc/refman/8.0/en/hash-joins.html