我目前在 mySQL 的“视图”中收到附加警告，知道这意味着啥吗？
Posted
技术标签:
【中文标题】我目前在 mySQL 的“视图”中收到附加警告，知道这意味着啥吗？【英文标题】：I am currently receiving the attached warning in a "View" in mySQL, any idea what this means?我目前在 mySQL 的“视图”中收到附加警告，知道这意味着什么吗？ 【发布时间】：2021-09-26 19:42:01 【问题描述】：Warning: #1292 Truncated incorrect DOUBLE value: "
错误：警告：#1292 截断不正确的 DOUBLE 值：“
我的“视图”的结构如下：
# | Name | Type
1 id1 varchar(255)
2 id2 varchar(255)
3 category1 varchar(255)
4 category2 varchar(255)
5 category3 varchar(255)
6 id2name varchar(255)
7 date1 date
8 date2 date
9 name1 varchar(255)
10 name2 varchar(255)
11 name3 varchar(255)
12 category4 varchar(255)
13 level1 int(11)
14 level2 int(11)
15 levelflex varchar(255)
16 category5 varchar(255)
17 category6 varchar(255)
18 category7 varchar(21)
SQL 代码：
SELECT
`x`.`id1` AS `id1`,
`x`.`id2` AS `id2`,
`x`.`category1` AS `category1`,
`x`.`category2` AS `category2`,
`x`.`category3` AS `category3`,
`x`.`id2name` AS `id2name`,
`x`.`date1` AS `date1`,
`x`.`date2` AS `date2`,
`x`.`name1` AS `name1`,
`x`.`name2` AS `name2`,
`x`.`name3` AS `name3`,
`x`.`category4` AS `category4`,
`x`.`level1` AS `level1`,
`x`.`level2` AS `level2`,
`x`.`levelflex` AS `levelflex`,
`x`.`category5` AS `category5`,
`x`.`category6` AS `category6`,
`x`.`category7` AS `category7`FROM
(
(
SELECT
`r`.`id1` AS `id1`,
`d`.`id2` AS `id2`,
`m`.`subcategory` AS `category1`,
`n`.`subcategory` AS `category2`,
`r`.`category3` AS `category3`,
`d`.`id2name` AS `id2name`,
`r`.`date1` AS `date1`,
`d`.`date2` AS `date2`,
`r`.`name1` AS `name1`,
`r`.`name2` AS `name2`,
`r`.`name3` AS `name3`,
`r`.`category4` AS `category4`,
`r`.`level1` AS `level1`,
`d`.`level1` AS `level2`,
`d`.`level1flex` AS `levelflex`,
`r`.`category5` AS `category5`,
`r`.`category6` AS `category6`,
CAST(
rank() over(
PARTITION BY `r`.`id1`
ORDER BY CASE WHEN
`m`.`subcategory` = `n`.`subcategory` AND `r`.`level1` = `d`.`level1` THEN 1 WHEN `m`.`subcategory` = `n`.`subcategory` AND `r`.`level1` = `d`.`levelflex` THEN 2 WHEN `m`.`subcategory` <> `n`.`subcategory` AND `r`.`level1` = `d`.`level1` THEN 3 WHEN `m`.`subcategory` <> `n`.`subcategory` AND `r`.`level1` = `d`.`levelflex` THEN 4
END,
`d`.`date2`,
`r`.`id1`
) AS CHAR CHARSET utf8mb4) AS `category7`FROM
(
(
(
(
(
(
(
(
`Server`.`table1` `d`
JOIN `Server`.`table2` `r`
ON
(
`d`.`category4` = `r`.`category4` AND(
`d`.`level1` = `r`.`level1` AND `r`.`level1` <> 12 AND `r`.`level1` <> 13 OR SUBSTRING_INDEX(`d`.`levelflex`, ',', 1) = `r`.`level1` AND `r`.`level1` <> 12 AND `r`.`level1` <> 13 OR SUBSTRING_INDEX(`d`.`levelflex`, ',', -1) = `r`.`level1` AND `r`.`level1` <> 12 AND `r`.`level1` <> 13
)
)
)
JOIN `Server`.`table3` `p`
ON
(
`d`.`id2name` = `p`.`newid2name` OR `d`.`id2name` = `p`.`oldid2name`
)
)
JOIN `Server`.`table4` `m`
ON
(
`p`.`newid2name` = `m`.`id2name` OR `p`.`oldid2name` = `m`.`id2name`
)
)
JOIN `Server`.`table5` `h`
ON
(`r`.`category3` = `h`.`category3`)
)
JOIN `Server`.`table6` `f`
ON
(`r`.`id1` = `f`.`id1`)
)
LEFT JOIN `Server`.`table7` `e`
ON
(`f`.`id1` = `e`.`id1`)
)
JOIN `Server`.`table8` `n`
ON
(
(
`r`.`date1` >= `h`.`subdate1` AND `e`.`subdate1` = `h`.`subdate1` OR `f`.`subdate2` = `h`.`subdate1` AND `r`.`date1` < `h`.`subdate1`
) AND `h`.`id2name` = `n`.`id2name`
)
)
LEFT JOIN `Server`.`table9` `z`
ON
(
`d`.`id2` = `z`.`id2` AND `r`.`id1` = `z`.`id1`
)
)
WHERE
`z`.`id2` IS NULL
) `x`LEFT JOIN `Server`.`table10` `y`ON
(
`x`.`id2` = `y`.`id2` AND `x`.`id1` = `y`.`id1`
)
)WHERE
`y`.`id2` IS NULL
总共 44 行的查询也需要大约 15 秒或更长时间。有什么办法可以优化吗？
我也在尝试将数据从视图传输到表。对此有何建议？
显示创建表：
| View | Create View | character_set_client | collation_connection |
|---|---|---|---|
| sampleview | CREATE ALGORITHM=UNDEFINED DEFINER=dbadmin@%SQL SECURITY DEFINER VIEWsampleviewAS selectx.id1 AS id1,x.`r... |
utf8mb4 | utfmb4_unicode_ci |
解释选择...：
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
|---|---|---|---|---|---|---|---|---|---|
| 1 | PRIMARY | derived2 | ALL | NULL | NULL | NULL | NULL | 251497948917600 | |
| 1 | PRIMARY | y | ALL | NULL | NULL | NULL | NULL | 1 | Using where; Using join buffer(flat, BNL join) |
| 2 | DERIVED | d | ALL | NULL | NULL | NULL | NULL | 106 | Using temporary |
| 2 | DERIVED | f | ALL | PRIMARY | NULL | NULL | NULL | 115 | Using join buffer(flat,BNL join) |
| 2 | DERIVED | e | eq_ref | PRIMARY | PRIMARY | 767 | Server.f.id1 | 1 | |
| 2 | DERIVED | r | eq_ref | PRIMARY | PRIMARY | 767 | Server.f.id1 | 1 | Using where |
| 2 | DERIVED | z | ALL | NULL | NULL | NULL | NULL | 1 | Using where; Using join buffer (flat, BNL join) |
| 2 | DERIVED | p | ALL | NULL | NULL | NULL | NULL | 140 | Using where; Using join buffer (incremental, BNL join) |
| 2 | DERIVED | m | ALL | NULL | NULL | NULL | NULL | 198 | Using where; Using join buffer (incremental, BNL join) |
| 2 | DERIVED | n | ALL | NULL | NULL | NULL | NULL | 198 | Using join buffer (flat, BNL join) |
| 2 | DERIVED | h | ALL | NULL | NULL | NULL | NULL | 3759 | Using where; Using join buffer (incremental, BNL join) |
【问题讨论】：
编辑以包含完整的 SQL，以及作为文本的完整错误消息。 【参考方案1】：有 251 个 trillion 行，我怀疑该查询包含意外的“交叉连接”。在VIEW 和VIEW 的使用中检查所有JOIN。
另外，请注意AND 和OR 之间的优先规则。
OR 通常是性能杀手。
（同时，我没有看到关于错误 1292 的任何线索。）
其中一些复合索引可能会有所帮助：
y: INDEX(id2, id1)
r: INDEX(category4)
d: INDEX(category4)
m: INDEX(id2name, subcategory)
n: INDEX(id2name, subcategory)
p: INDEX(newid2name, oldid2name)
h: INDEX(category3, subdate1, id2name)
f: INDEX(id1, subdate2)
e: INDEX(id1, subdate1)
z: INDEX(id2, id1)
部分性能优势在于它们中的大多数是“覆盖”。如需进一步讨论，请为基础表提供SHOW CREATE TABLE 和EXPLAIN SELECT...
【讨论】：
谢谢，瑞克。我添加了 SHOW CREATE TABLE 和 EXPLAIN SELECT... 我又添加了一些。 我会调查的，瑞克。非常感谢您的帮助。我目前无法投票，但我会投票给你。以上是关于我目前在 mySQL 的“视图”中收到附加警告，知道这意味着啥吗？的主要内容，如果未能解决你的问题，请参考以下文章
警告：mysql_query() 期望参数 2 是资源 [重复]