全表扫描

查询sql语句：
select Name ,Id, count(*)  from test where sesTime is not null group by Name order by count(*) desc limit 15;
############################
查看sql语句的执行计划：
desc select Name ,Id, count(*) from test where sesTime is not null group by Name order by count(*) desc limit 15;+----+-------------+----------------+------------+------+---------------+------+---------+------+---------+----------+----------------------------------------------+
| id | select_type | table          | partitions | type | possible_keys | key  | key_len | ref  | rows    | filtered | Extra                                        |
+----+-------------+----------------+------------+------+---------------+------+---------+------+---------+----------+----------------------------------------------+
|  1 | SIMPLE      | test | NULL       | ALL  | NULL          | NULL | NULL    | NULL | 1305648 |    90.00 | Using where; Using temporary; Using filesort |
+----+-------------+----------------+------------+------+---------------+------+---------+------+---------+----------+----------------------------------------------+
1 row in set, 1 warning (0.00 sec)

Thu Oct 10 18:34:36 2019

显然，这是一个全表扫描语句，
##########################
查看该表的数据量：
select count(*) from test;
+----------+
| count(*) |
+----------+
|  1373833 |
+----------+
1 row in set (0.48 sec)

Thu Oct 10 18:45:10 2019
显然，该表数据量有140万条记录左右。
############################
查看该表结构：
show create table mbi_order_tempG;
*************************** 1. row ***************************

Thu Oct 10 18:48:10 2019
从这个表结构语句可知：分组字段fiveGradeName没有索引，因此整个语句都是全表扫描
#######
#######

xxx@igoodful(yyy) > show create table job G
*************************** 1. row ***************************
       Table: job_stat
Create Table: CREATE TABLE `job_stat` (
  `id` bigint(20) unsigned NOT NULL AUTO_INCREMENT,
  `pdl` varchar(255) COLLATE utf8_unicode_ci NOT NULL COMMENT '产品线',
  `tags` varchar(255) COLLATE utf8_unicode_ci NOT NULL COMMENT 'j名称',
  `cluster` varchar(64) COLLATE utf8_unicode_ci NOT NULL COMMENT '所在集群',
  `cpus` float NOT NULL COMMENT 'cpu模版数量',
  `mem` float NOT NULL COMMENT 'mem模板数量',
  `disk` float NOT NULL COMMENT 'disk模板数量',
  `all_cpus` float NOT NULL COMMENT 'cpu总数量',
  `all_mem` float NOT NULL COMMENT 'mem总数量',
  `all_disk` float NOT NULL COMMENT 'disk总数量',
  `used_cpus` float NOT NULL COMMENT 'cpu使用数量',
  `used_mem` float NOT NULL COMMENT 'mem使用数量',
  `used_disk` float NOT NULL COMMENT 'disk使用数量',
  `running` int(11) NOT NULL COMMENT 'task running 数量',
  `instances` int(11) DEFAULT NULL COMMENT 'instances数量',
  `timestamp` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '时间',
  PRIMARY KEY (`id`),
  KEY `idx_job_ts` (`tags`,`timestamp`),
  KEY `idx_pdl` (`pdl`),
  KEY `idx_ts` (`timestamp`)
) ENGINE=InnoDB AUTO_INCREMENT=75982981 DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci
1 row in set (0.00 sec)

xxx@igoodful(yyy) > select count(*) from job; 

+----------+
| count(*) |
+----------+
| 62680162 |
+----------+
1 row in set (10.11 sec)




#################################################

下面就是一个全表扫描语句：

原因：like使用了前导%，导致索引失效，进而全表扫描。该语句执行差不多需要2分钟。

#################################################

xxx@igoodful(yyy) > select * from job where tags like "%service.browser-score-service_cluster.c3_pdl.browser_owt.miui_cop.xiaomi%";