“明明给字段建了索引，可带函数的查询还是走全表扫描，数据量大了直接卡到超时……”
如果你在使用 OceanBase 或 MySQL 时遇到过这种问题，大概率是没用到函数索引！当查询条件包含函数计算（比如LOWER(name)、MD5(customer_id)）时，普通索引完全失效，而函数索引能直接基于计算结果建立索引，让查询效率飙升10倍+！
今天这篇干货，从定义、用法、避坑到实战场景，带你彻底掌握 OceanBase 函数索引的核心玩法～
1. 函数索引的定义
函数索引是基于表达式或函数计算结果创建的索引，而非直接基于列值。
参考文档：MySQL 8.0 参考手册
2. 函数索引的作用
- 当查询条件包含函数时，普通索引无法使用，优化函数表达式查询；
- 对函数计算结果建立索引可以避免全表扫描；
- 对于包含计算的查询条件能显著提高效率；
2.1 函数索引的创建
ps：普通索引为 [索引名](列名)，函数索引为 [索引名]((列名));两层括号后开始填写函数表达式；
创建函数索引时，会向主表添加一个隐藏的虚拟生成列，定义为函数索引的索引键，然后再将虚拟生成列的值存储到索引表中。
/* 建表时创建*/
CREATE TABLE tst_tb (col1 int, col2 int, [UNIQUE] {KEY / INDEX} [索引名]((用户函数(col1)));
/* 使用 create index 创建 */
CREATE [UNIQUE] INDEX tst_tb_func_idx2 ON tst_tb((col1+3));
/* 使用 alter table 创建 */
ALTER TABLE tst_tb ADD [UNIQUE] {INDEX | KEY} [索引名] ((用户函数(col1)));
2.2 函数索引的查询
/*
示例1：对名字的小写形式创建索引,使用与索引定义完全相同的表达式查询
*/
CREATE INDEX idx_lower_name ON tst_tb ((LOWER(name)));
SELECT * FROM employees WHERE LOWER(name) = 'John';
/*
示例2：对JSON字段的特定路径创建索引
从 JSON 文档中指定的路径返回数据
*/
CREATE INDEX idx_json_extract ON tst_tb ((JSON_EXTRACT(person, '$.weight')));
SELECT * FROM tst_tb WHERE JSON_EXTRACT(specs, '$.weight') > 10;
3.函数索引的使用限制
> 1. 非确定性的系统函数禁止被用于创建函数索引;
> 2. 子查询、参数、变量、存储过程、自定义函数 function 不允许在函数索引上使用；
> 3. 函数索引的表达式结果受限于字段长度；
> 4. 函数索引无法作为主键使用；
> 5. 空间索引和全文索引不能定义为函数索引；
> 6. 函数索引不允许在外键中使用；
> 7. 某字段用于函数索引，删列前必须删除该索引，否则报错；
/* 1. 使用非确定性函数*/
create index tst_01 on tst_tb((now()));
失败原因：ErrorCode = 3758, SQLState = HY000, Details = Expression of functional index contains a disallowed function.
/*在创建函数索引时，使用了不支持的函数，索引函数使用 SYSDATE 或用户环境*/
create index current_time_fun_idx on tst_tb((current_date()));
失败原因：ErrorCode = 3758, SQLState = HY000, Details = Expression of functional index contains a disallowed function.
/* 2.表达式过于复杂，函数索引中包括了子查询*/
CREATE INDEX tst_02 ON tst_tb ((SELECT MAX(id) FROM tst_tb));
失败原因：ErrorCode = 1064, SQLState = 42000, Details = You have an error in your SQL syntax; check the manual that corresponds to your OceanBase version for the right syntax to use near 'SELECT MAX(id) FROM tst_tb))' at line 1
/* 3.表达式结果太长,REPEAT 重复多次改字段字符串拼接*/
CREATE INDEX idx_invalid ON tst_tb ((REPEAT(customer_id, 100)));
失败原因：ErrorCode = 3757, SQLState = HY000, Details = Cannot create a functional index on an expression that returns a BLOB or TEXT. Please consider using CAST.
/* 在创建函数索引时，使用了返回 BLOB 或 TEXT 类型的表达式，而 BLOB 或 TEXT 类型的表达式不能用于创建函数索引 */
/* 4.函数索引无法作为主键创建*/
CREATE TABLE `tst_tb` (
`id` bigint(20) unsigned NOT NULL AUTO_INCREMENT COMMENT '主键id，自动增长',
`customer_id` varchar(50) DEFAULT NULL COMMENT '客户ID',
`create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
`update_time` datetime DEFAULT NULL ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
`product` varchar(10) DEFAULT NULL COMMENT '产品类型',
PRIMARY KEY `pri_cust` ((lower(md5(concat(upper(customer_id), 'oceanbase'))))),
KEY `idx_create_time` (`create_time`) BLOCK_SIZE 16384 LOCAL,
KEY `ob_idx_customer_id` (`customer_id`) BLOCK_SIZE 16384 LOCAL
);
失败原因：ErrorCode = 1064, SQLState = 42000, Details = You have an error in your SQL syntax; check the manual that corresponds to your OceanBase version for the right syntax to use near '(lower(md5(concat(upper(customer_id), 'oceanbase'))))), KEY `idx_create_time` ' at line 7
/* 5.空间索引和全文索引不能定义为函数索引*/
CREATE TABLE tbl_func (c1 int, c2 int, index tbl1_func_idx1 ((c1+1)), SPATIAL KEY ((c1+c2)));
失败原因：ErrorCode = 3760, SQLState = HY000, Details = Spatial functional index is not supported.
/* 不支持空间索引套用函数索引*/
CREATE TABLE tbl_func (c1 int, c2 varchar(20), index tbl1_func_idx1 ((c1+1)), FULLTEXT INDEX full_idx_func ((SUBSTR(c2,1,2))) WITH PARSER ngram PARSER_PROPERTIES=(ngram_token_size=3));
失败原因：ErrorCode = 1235, SQLState = 0A000, Details = Fulltext index on generated column is not supported
/* 不支持全文索引套用函数索引*/
/* 6.函数索引不允许在外键中使用*/
CREATE TABLE products (
product_id INT PRIMARY KEY,
product_code VARCHAR(20) NOT NULL,
product_name VARCHAR(100) NOT NULL,
INDEX idx_lower_product_code ((LOWER(product_code)))
);
CREATE TABLE orders (
order_id INT PRIMARY KEY,
product_code_ref VARCHAR(20),
CONSTRAINT fk_order_product FOREIGN KEY (product_code_ref) -- 创建外键引用函数索引
REFERENCES products (LOWER(product_code))
);
失败原因：ErrorCode = 1064, SQLState = 42000, Details = You have an error in your SQL syntax; check the manual that corresponds to your OceanBase version for the right syntax to use near '(product_code)) )' at line 7
/* 7. 某字段用于函数索引，删列前必须删除该索引，否则报错*/
ALTER TABLE products DROP COLUMN product_code;
失败原因：ErrorCode = 3837, SQLState = HY000, Details = Column 'product_code' has a functional index dependency and cannot be dropped or renamed.
参考文档：
腾讯云：mysql函数索引_MySQL 函数索引 (Functional indexes)
函数索引支持的系统函数列表
函数索引不支持的系统函数列表
MySQL 函数索引测试分析
4. 无法创建函数索引的替代方案
/*表定义语句*/
CREATE TABLE `tst_tb` (
`id` bigint(20) unsigned NOT NULL AUTO_INCREMENT COMMENT '主键id，自动增长',
`customer_id` varchar(50) DEFAULT NULL COMMENT '客户ID',
`create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
`update_time` datetime DEFAULT NULL ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
`product` varchar(10) DEFAULT NULL COMMENT '产品类型',
`name` varchar(100) DEFAULT NULL,
PRIMARY KEY (`id`),
KEY `idx_create_time` (`create_time`) BLOCK_SIZE 16384 LOCAL,
KEY `ob_idx_customer_id` (`customer_id`) BLOCK_SIZE 16384 LOCAL
);
explain extended select * from tst_tb where name='Product One';
=================================================
|ID|OPERATOR |NAME |EST.ROWS|EST.TIME(us)|
-------------------------------------------------
|0 |TABLE FULL SCAN|tst_tb|1 |3 |
=================================================
Outputs & filters:
-------------------------------------
0 - output([tst_tb.id(0x7ff8f1823480)], [tst_tb.customer_id(0x7ff8f18238d0)], [tst_tb.create_time(0x7ff8f1823d20)], [tst_tb.update_time(0x7ff8f1824170)],
[tst_tb.product(0x7ff8f18245c0)], [tst_tb.name(0x7ff8f1822d60)]), filter([tst_tb.name(0x7ff8f1822d60) = 'Product One'(0x7ff8f1822000)(0x7ff8f1822490)]), rowset=16
access([tst_tb.id(0x7ff8f1823480)], [tst_tb.name(0x7ff8f1822d60)], [tst_tb.customer_id(0x7ff8f18238d0)], [tst_tb.create_time(0x7ff8f1823d20)], [tst_tb.update_time(0x7ff8f1824170)],
[tst_tb.product(0x7ff8f18245c0)]), partitions(p0)
is_index_back=false, is_global_index=false, filter_before_indexback[false],
range_key([tst_tb.id(0x7ff8f1823480)]), range(MIN ; MAX)always true
/* 添加虚拟生成列 */
ALTER TABLE tst_tb ADD COLUMN lower_name VARCHAR(100)
GENERATED ALWAYS AS (LOWER(name)) STORED;
/* 在虚拟生成列上创建索引 */
CREATE INDEX idx_lower_name ON tst_tb (lower_name);
/* 执行计划 */
explain extended select * from tst_tb where lower_name='Product One';
==================================================================
|ID|OPERATOR |NAME |EST.ROWS|EST.TIME(us)|
------------------------------------------------------------------
|0 |TABLE RANGE SCAN|tst_tb(idx_lower_name)|1 |8 |
==================================================================
Outputs & filters:
-------------------------------------
0 - output([tst_tb.id(0x7ff8bd22d120)], [tst_tb.customer_id(0x7ff8bd22d570)], [tst_tb.create_time(0x7ff8bd22d9c0)], [tst_tb.update_time(0x7ff8bd22de10)],
[tst_tb.product(0x7ff8bd22e260)], [tst_tb.name(0x7ff8bd223840)], [tst_tb.lower_name(0x7ff8bd221bc0)]), filter(nil), rowset=16
access([tst_tb.id(0x7ff8bd22d120)], [tst_tb.name(0x7ff8bd223840)], [tst_tb.lower_name(0x7ff8bd221bc0)], [tst_tb.customer_id(0x7ff8bd22d570)], [tst_tb.create_time(0x7ff8bd22d9c0)],
[tst_tb.update_time(0x7ff8bd22de10)], [tst_tb.product(0x7ff8bd22e260)]), partitions(p0)
is_index_back=true, is_global_index=false,
range_key([tst_tb.lower_name(0x7ff8bd221bc0)], [tst_tb.id(0x7ff8bd22d120)]), range(Product One,MIN ; Product One,MAX),
range_cond([tst_tb.lower_name(0x7ff8bd221bc0) = 'Product One'(0x7ff8bd2eb070)(0x7ff8bd2ea650)])
表数据
未走索引
走索引
客户使用场景
场景一：给表字段添加函数索引进行优化
/*
A 表结构
B 表与A表结构一致，且B表的数据是异步更新
*/
CREATE TABLE `tst_ta` (
`id` bigint(20) unsigned NOT NULL AUTO_INCREMENT COMMENT '主键id，自动增长',
`customer_id` varchar(50) DEFAULT NULL COMMENT '客户ID',
`create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
`update_time` datetime DEFAULT NULL ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
`product` varchar(10) DEFAULT NULL COMMENT '产品类型',
PRIMARY KEY (`id`, `create_time`),
KEY `idx_create_time` (`create_time`) BLOCK_SIZE 16384 LOCAL,
KEY `ob_idx_customer_id` (`customer_id`) BLOCK_SIZE 16384 LOCAL
);
测试表模拟数据
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST001', '2023-01-15 10:30:00', 'A');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST002', '2023-01-16 11:45:00', 'B');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST003', '2023-01-17 09:15:00', 'C');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST004', '2023-01-18 14:20:00', 'A');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST005', '2023-01-19 16:30:00', 'B');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST006', '2023-01-20 10:10:00', 'D');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST007', '2023-01-21 13:25:00', 'A');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST008', '2023-01-22 15:40:00', 'C');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST009', '2023-01-23 08:50:00', 'B');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST010', '2023-01-24 17:00:00', 'D');
INSERT INTO `tst_tb` (`customer_id`, `product`) VALUES ('CUST011', 'E');
INSERT INTO `tst_tb` (`customer_id`, `product`) VALUES ('CUST012', 'A');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST013', '2023-01-27 12:15:00', 'B');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST014', '2023-01-28 14:30:00', 'C');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST015', '2023-01-29 09:45:00', 'D');
INSERT INTO `tst_tb` (`customer_id`, `product`) VALUES ('CUST016', 'A');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST017', '2023-01-31 11:20:00', 'B');
INSERT INTO `tst_tb` (`customer_id`, `product`) VALUES ('CUST018', 'E');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST019', '2023-02-02 16:10:00', 'C');
INSERT INTO `tst_tb` (`customer_id`, `update_time`, `product`) VALUES ('CUST020', '2023-02-03 10:25:00', 'D');
tst_ta.customer_id 加密后的表数据
tst_tb.customer_id 加密前的表数据
/*
将 customer_id 字段的值统一大写后与'oceanbase'拼接，然后经过 md5 算法加密后统一小写后记录
*/
CREATE VIEW `tst_view` AS
( select id,customer_id,create_time,update_time,product from tst_ta)
union all
(select id,
lower(md5(concat(upper(customer_id), 'oceanbase'))) AS customer_id,
create_time,update_time,product from tst_tb
);
未走索引
/*
CREATE TABLE `tst_ta` (
`id` bigint(20) unsigned NOT NULL AUTO_INCREMENT COMMENT '主键id，自动增长',
`customer_id` varchar(50) DEFAULT NULL COMMENT '客户ID',
`create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
`update_time` datetime DEFAULT NULL ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
`product` varchar(10) DEFAULT NULL COMMENT '产品类型',
PRIMARY KEY (`id`, `create_time`),
KEY `idx_create_time` (`create_time`) BLOCK_SIZE 16384 LOCAL,
KEY `ob_idx_customer_id` (`customer_id`) BLOCK_SIZE 16384 LOCAL
);
*/
explain extended select * from tst_view where customer_id='366330e31189acd28639de9dd7b29cb1';
========================================================================
|ID|OPERATOR |NAME |EST.ROWS|EST.TIME(us)|
------------------------------------------------------------------------
|0 |UNION ALL | |2 |11 |
|1 |├─TABLE RANGE SCAN|tst_ta(ob_idx_customer_id)|1 |7 |
|2 |└─TABLE FULL SCAN |tst_tb |1 |4 |
========================================================================
Outputs & filters:
-------------------------------------
0 - output([UNION([1])(0x7ff9ce07efa0)], [UNION([2])(0x7ff9ce07f360)], [UNION([3])(0x7ff9ce07f720)], [UNION([4])(0x7ff9ce07fae0)], [UNION([5])(0x7ff9ce07fea0)]), filter(nil), rowset=16
1 - output([tst_ta.id(0x7ff9ce064850)], [tst_ta.customer_id(0x7ff9ce065120)], [tst_ta.create_time(0x7ff9ce0659f0)], [tst_ta.update_time(0x7ff9ce0662c0)],
[tst_ta.product(0x7ff9ce066b90)]), filter(nil), rowset=16
access([tst_ta.__pk_increment(0x7ff9ce084030)], [tst_ta.id(0x7ff9ce064850)], [tst_ta.customer_id(0x7ff9ce065120)], [tst_ta.create_time(0x7ff9ce0659f0)],
[tst_ta.update_time(0x7ff9ce0662c0)], [tst_ta.product(0x7ff9ce066b90)]), partitions(p0)
is_index_back=true, is_global_index=false,
range_key([tst_ta.customer_id(0x7ff9ce065120)], [tst_ta.__pk_increment(0x7ff9ce084030)]), range(366330e31189acd28639de9dd7b29cb1,MIN ; 366330e31189acd28639de9dd7b29cb1,
MAX),
range_cond([tst_ta.customer_id(0x7ff9ce065120) = '366330e31189acd28639de9dd7b29cb1'(0x7ff8f0484a00)(0x7ff8f0483fe0)])
2 - output([tst_tb.id(0x7ff9ce0797e0)], [lower(md5(concat(upper(tst_tb.customer_id(0x7ff9ce07ce60))(0x7ff9ce07b6b0), 'oceanbase')(0x7ff9ce07ade0))(0x7ff9ce07a510))(0x7ff9ce079c40)],
[tst_tb.create_time(0x7ff9ce07d870)], [tst_tb.update_time(0x7ff9ce07e140)], [tst_tb.product(0x7ff9ce07ea10)]), filter([lower(md5(concat(upper(tst_tb.customer_id(0x7ff9ce07ce60))(0x7ff9ce07b6b0),
'oceanbase')(0x7ff9ce07ade0))(0x7ff9ce07a510))(0x7ff9ce079c40) = '366330e31189acd28639de9dd7b29cb1'(0x7ff9ce0813f0)(0x7ff9ce0991d0)]), rowset=16
access([tst_tb.id(0x7ff9ce0797e0)], [tst_tb.create_time(0x7ff9ce07d870)], [tst_tb.customer_id(0x7ff9ce07ce60)], [tst_tb.update_time(0x7ff9ce07e140)],
[tst_tb.product(0x7ff9ce07ea10)]), partitions(p0)
is_index_back=false, is_global_index=false, filter_before_indexback[false],
range_key([tst_tb.id(0x7ff9ce0797e0)], [tst_tb.create_time(0x7ff9ce07d870)]), range(MIN,MIN ; MAX,MAX)always true
走索引
/*添加索引 1*/
create index tst_tb_fun_idx on tst_tb((lower(md5(concat(upper(customer_id), 'oceanbase')))));
/*
执行 select * from tst_view where customer_id='366330e31189acd28639de9dd7b29cb1';
相当于
select id,customer_id,create_time,update_time,product from tst_ta where customer_id='366330e31189acd28639de9dd7b29cb1'
union all
select id,
lower(md5(concat(upper(customer_id), 'oceanbase'))) AS customer_id,
create_time,update_time,product from tst_tb where lower(md5(concat(upper(customer_id), 'oceanbase')))='366330e31189acd28639de9dd7b29cb1';
而不是 select * from tst_tb where customer_id='366330e31189acd28639de9dd7b29cb1';
*/
/* 执行计划 */
explain extended select * from tst_view where customer_id='366330e31189acd28639de9dd7b29cb1';
============================================================================
|ID|OPERATOR |NAME |EST.ROWS|EST.TIME(us)|
----------------------------------------------------------------------------
|0 |EXCHANGE IN REMOTE | |2 |22 |
|1 |└─EXCHANGE OUT REMOTE | |2 |19 |
|2 | └─UNION ALL | |2 |14 |
|3 | ├─TABLE RANGE SCAN|tst_ta(ob_idx_customer_id)|1 |7 |
|4 | └─TABLE RANGE SCAN|tst_tb(tst_tb_fun_idx) |1 |8 |
============================================================================
Outputs & filters:
-------------------------------------
0 - output([UNION([1])(0x7fd63d894110)], [UNION([2])(0x7fd63d8944d0)], [UNION([3])(0x7fd63d894890)], [UNION([4])(0x7fd63d894c50)], [UNION([5])(0x7fd63d895010)]), filter(nil)
1 - output([UNION([1])(0x7fd63d894110)], [UNION([2])(0x7fd63d8944d0)], [UNION([3])(0x7fd63d894890)], [UNION([4])(0x7fd63d894c50)], [UNION([5])(0x7fd63d895010)]), filter(nil)
2 - output([UNION([1])(0x7fd63d894110)], [UNION([2])(0x7fd63d8944d0)], [UNION([3])(0x7fd63d894890)], [UNION([4])(0x7fd63d894c50)], [UNION([5])(0x7fd63d895010)]), filter(nil), rowset=16
3 - output([tst_ta.id(0x7fd63d864850)], [tst_ta.customer_id(0x7fd63d865120)], [tst_ta.create_time(0x7fd63d8659f0)], [tst_ta.update_time(0x7fd63d8662c0)],
[tst_ta.product(0x7fd63d866b90)]), filter(nil), rowset=16
access([tst_ta.__pk_increment(0x7fd63d8a5180)], [tst_ta.id(0x7fd63d864850)], [tst_ta.customer_id(0x7fd63d865120)], [tst_ta.create_time(0x7fd63d8659f0)],
[tst_ta.update_time(0x7fd63d8662c0)], [tst_ta.product(0x7fd63d866b90)]), partitions(p0)
is_index_back=true, is_global_index=false,
range_key([tst_ta.customer_id(0x7fd63d865120)], [tst_ta.__pk_increment(0x7fd63d8a5180)]), range(366330e31189acd28639de9dd7b29cb1,MIN ; 366330e31189acd28639de9dd7b29cb1,
MAX),
range_cond([tst_ta.customer_id(0x7fd63d865120) = '366330e31189acd28639de9dd7b29cb1'(0x7fd76aeda000)(0x7fd76aed95e0)])
4 - output([tst_tb.id(0x7fd63d886840)], [lower(md5(concat(upper(tst_tb.customer_id(0x7fd63d87ddf0))(0x7fd63d888710), 'oceanbase')(0x7fd63d887e40))(0x7fd63d887570))(0x7fd63d886ca0)],
[tst_tb.create_time(0x7fd63d88a480)], [tst_tb.update_time(0x7fd63d88ad50)], [tst_tb.product(0x7fd63d88b620)]), filter(nil), rowset=16
access([tst_tb.id(0x7fd63d886840)], [tst_tb.create_time(0x7fd63d88a480)], [tst_tb.customer_id(0x7fd63d87ddf0)], [tst_tb.update_time(0x7fd63d88ad50)],
[tst_tb.product(0x7fd63d88b620)]), partitions(p0)
is_index_back=true, is_global_index=false,
range_key([tst_tb.SYS_NC21$(0x7fd63d879390)], [tst_tb.id(0x7fd63d886840)], [tst_tb.create_time(0x7fd63d88a480)]), range(366330e31189acd28639de9dd7b29cb1,
MIN,MIN ; 366330e31189acd28639de9dd7b29cb1,MAX,MAX),
range_cond([tst_tb.SYS_NC21$(0x7fd63d879390) = '366330e31189acd28639de9dd7b29cb1'(0x7fd76aedaeb0)(0x7fd76aeda490)])
tst_tb.customer_id='...'的执行计划
tst_tb.lower(md5(concat(upper(customer_id), 'oceanbase')))='...'的执行计划
以上就是 OceanBase 函数索引的核心内容， 如果你在操作中遇到问题，或是有独到的优化建议，欢迎在评论区留言分享，后续我们还会分享更多OceanBase / MySQL 性能优化秘籍，下次见~
