如何通过多个连接加速 MYSQL 查询？

Question

寻求帮助以加快此查询：

SELECT tickets.ticketid, datenew, products.name 
FROM tickets 
INNER JOIN ticketlines ON tickets.id = ticketlines.ticket 
INNER JOIN products ON ticketlines.product = products.id 
INNER JOIN receipts ON receipts.id = tickets.id 
WHERE (category !='feb765ef-c8a8-4fa2-969c-90f67fe6b3be' AND category!='888f4893-f300-43b5-9933-d549ade744e0' AND category !='8f2031e8-64a4-4abf-8175-3d2bedd9f950' AND category !='ca370ced-1c3b-434c-905e-ec1bc709543b' AND category !='f92ff0ac-fa11-4a5f-a3dd-e0d9ed9c171a' AND category !='445e8605-1cd9-4714-b3fd-7389ac29c206' and category !='05143c54-8a7e-4ce2-97cc-f84f9cf41395' AND category !='8c78afea-b9e2-44cf-b497-c384045b3202' AND category !='95919f7f-ff2e-4aa1-8110-ef63c022c01b' AND category !='f4f88b05-38a1-4956-9182-4c04a0808df7') AND datedone IS NULL 
ORDER BY ticketid

这是我用来提取订单显示数据的收银机数据库。 结构是收据具有主 ID、timein 和 timedone(DATEDONE) 时间戳。 票证表的票证 ID 与收据 ID 相同。 Ticketlines 表是收据/票上的行项目。 然后 Products 表具有产品的人类可读定义。

查询是拉取所有未完成的项目，即DATEDONE为null，并显示已订购的项目。

Ticket 和 Receipt 表有 15K 行，Ticketlines 有大约 20K 行，Products 有 1500 行。

相当小的数据。 但是这个查询需要超过 20 秒。 我想因为我对所有东西都使用主键 ID，所以我不需要索引任何东西，但我是一个完全的菜鸟，所以我很高兴被告知我错了。 我将不胜感激任何帮助，并可以提供任何进一步的细节。 谢谢！

编辑：

根据评论，我正在展示表结构。 对于格式化噩梦，我很抱歉，我不熟悉这个平台并且不确定如何使其更具可读性:-(

| receipts | CREATE TABLE `receipts` (
  `ID` varchar(255) NOT NULL,
  `MONEY` varchar(255) NOT NULL,
  `DATENEW` datetime NOT NULL,
  `ATTRIBUTES` mediumblob,
  `PERSON` varchar(255) DEFAULT NULL,
  `DATEDONE` datetime DEFAULT NULL,
  PRIMARY KEY (`ID`),
  KEY `RECEIPTS_FK_MONEY` (`MONEY`),
  KEY `RECEIPTS_INX_1` (`DATENEW`),
  CONSTRAINT `RECEIPTS_FK_MONEY` FOREIGN KEY (`MONEY`) REFERENCES `closedcash` (`MONEY`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 |


| tickets | CREATE TABLE `tickets` (
  `ID` varchar(255) NOT NULL,
  `TICKETTYPE` int(11) NOT NULL DEFAULT '0',
  `TICKETID` int(11) NOT NULL,
  `PERSON` varchar(255) NOT NULL,
  `CUSTOMER` varchar(255) DEFAULT NULL,
  `STATUS` int(11) NOT NULL DEFAULT '0',
  `DONE` varchar(255) DEFAULT NULL,
  PRIMARY KEY (`ID`),
  KEY `TICKETS_FK_2` (`PERSON`),
  KEY `TICKETS_CUSTOMERS_FK` (`CUSTOMER`),
  KEY `TICKETS_TICKETID` (`TICKETTYPE`,`TICKETID`),
  CONSTRAINT `TICKETS_CUSTOMERS_FK` FOREIGN KEY (`CUSTOMER`) REFERENCES `customers` (`ID`),
  CONSTRAINT `TICKETS_FK_2` FOREIGN KEY (`PERSON`) REFERENCES `people` (`ID`),
  CONSTRAINT `TICKETS_FK_ID` FOREIGN KEY (`ID`) REFERENCES `receipts` (`ID`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 |


| ticketlines | CREATE TABLE `ticketlines` (
  `TICKET` varchar(255) NOT NULL,
  `LINE` int(11) NOT NULL,
  `PRODUCT` varchar(255) DEFAULT NULL,
  `ATTRIBUTESETINSTANCE_ID` varchar(255) DEFAULT NULL,
  `UNITS` double NOT NULL,
  `PRICE` double NOT NULL,
  `TAXID` varchar(255) NOT NULL,
  `ATTRIBUTES` mediumblob,
  PRIMARY KEY (`TICKET`,`LINE`),
  KEY `TICKETLINES_FK_2` (`PRODUCT`),
  KEY `TICKETLINES_ATTSETINST` (`ATTRIBUTESETINSTANCE_ID`),
  KEY `TICKETLINES_FK_3` (`TAXID`),
  CONSTRAINT `TICKETLINES_ATTSETINST` FOREIGN KEY (`ATTRIBUTESETINSTANCE_ID`) REFERENCES `attributesetinstance` (`ID`),
  CONSTRAINT `TICKETLINES_FK_2` FOREIGN KEY (`PRODUCT`) REFERENCES `products` (`ID`),
  CONSTRAINT `TICKETLINES_FK_3` FOREIGN KEY (`TAXID`) REFERENCES `taxes` (`ID`),
  CONSTRAINT `TICKETLINES_FK_TICKET` FOREIGN KEY (`TICKET`) REFERENCES `tickets` (`ID`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 |

| products | CREATE TABLE `products` (
  `ID` varchar(255) NOT NULL,
  `REFERENCE` varchar(255) NOT NULL,
  `CODE` varchar(255) NOT NULL,
  `CODETYPE` varchar(255) DEFAULT NULL,
  `NAME` varchar(255) NOT NULL,
  `PRICEBUY` double NOT NULL,
  `PRICESELL` double NOT NULL,
  `CATEGORY` varchar(255) NOT NULL,
  `TAXCAT` varchar(255) NOT NULL,
  `ATTRIBUTESET_ID` varchar(255) DEFAULT NULL,
  `STOCKCOST` double DEFAULT NULL,
  `STOCKVOLUME` double DEFAULT NULL,
  `IMAGE` mediumblob,
  `ISCOM` bit(1) NOT NULL DEFAULT b'0',
  `ISSCALE` bit(1) NOT NULL DEFAULT b'0',
  `ISKITCHEN` bit(1) NOT NULL DEFAULT b'0',
  `PRINTKB` bit(1) NOT NULL DEFAULT b'0',
  `SENDSTATUS` bit(1) NOT NULL DEFAULT b'0',
  `ISSERVICE` bit(1) NOT NULL DEFAULT b'0',
  `ATTRIBUTES` mediumblob,
  `DISPLAY` varchar(255) DEFAULT NULL,
  PRIMARY KEY (`ID`),
  UNIQUE KEY `PRODUCTS_INX_0` (`REFERENCE`),
  UNIQUE KEY `PRODUCTS_INX_1` (`CODE`),
  UNIQUE KEY `PRODUCTS_NAME_INX` (`NAME`),
  KEY `PRODUCTS_FK_1` (`CATEGORY`),
  KEY `PRODUCTS_TAXCAT_FK` (`TAXCAT`),
  KEY `PRODUCTS_ATTRSET_FK` (`ATTRIBUTESET_ID`),
  CONSTRAINT `PRODUCTS_ATTRSET_FK` FOREIGN KEY (`ATTRIBUTESET_ID`) REFERENCES `attributeset` (`ID`),
  CONSTRAINT `PRODUCTS_FK_1` FOREIGN KEY (`CATEGORY`) REFERENCES `categories` (`ID`),
  CONSTRAINT `PRODUCTS_TAXCAT_FK` FOREIGN KEY (`TAXCAT`) REFERENCES `taxcategories` (`ID`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 |

此外，这里是 EXPLAIN output：

| id | select_type | table       | partitions | type   | possible_keys            | key     | key_len | ref                                | rows  | filtered | Extra                                        |
+----+-------------+-------------+------------+--------+--------------------------+---------+---------+------------------------------------+-------+----------+----------------------------------------------+
|  1 | SIMPLE      | receipts    | NULL       | ALL    | PRIMARY                  | NULL    | NULL    | NULL                               | 14624 |    10.00 | Using where; Using temporary; Using filesort |
|  1 | SIMPLE      | tickets     | NULL       | eq_ref | PRIMARY                  | PRIMARY | 767     | receipts.ID         |     1 |   100.00 | NULL                                         |
|  1 | SIMPLE      | ticketlines | NULL       | ref    | PRIMARY,TICKETLINES_FK_2 | PRIMARY | 767     | receipts.ID         |     1 |   100.00 | Using where                                  |
|  1 | SIMPLE      | products    | NULL       | eq_ref | PRIMARY,PRODUCTS_FK_1    | PRIMARY | 767     | ticketlines.PRODUCT |     1 |    97.97 | Using where                                  |
+----+-------------+-------------+------------+--------+--------------------------+---------+---------+------------------------------------+-------+----------+----------------------------------------------+
4 rows in set, 1 warning (0.04 sec)

Answer 1

尝试将所有列的数据类型更改为所需的最少类型。 在 WHERE 中的列上添加索引。 在 ticketlines.ticket 上添加索引。

Answer 2

我最终从每个链接表中删除了 10K 行，现在查询运行时间为 2 秒。 我想如果没有更多建议，我将每年清理一次旧票。