Question

表详情 - ＆gt;

CREATE TABLE `redeem_vouchers` (
  `voucherid` int(11) NOT NULL,
  `voucheramount` decimal(18,2) DEFAULT NULL,
  `redeemedto` int(11) DEFAULT NULL,
  `redeemedon` datetime DEFAULT NULL,
  `transid` varchar(100) DEFAULT NULL,
  `paymentconfirmed` tinyint(1) NOT NULL DEFAULT '0',
  `confirmedby` int(11) DEFAULT NULL,
  `confirmedon` datetime DEFAULT NULL,
  `redeemedby` int(11) DEFAULT '0',
  PRIMARY KEY (`voucherid`),
  KEY `NewIndex2` (`redeemedto`),
  KEY `NewIndex3` (`voucheramount`),
  FULLTEXT KEY `NewIndex1` (`transid`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

查询 - ＆GT;

SELECT DISTINCT(transid),
    (SELECT (COUNT(voucheramount)*16) FROM redeem_vouchers WHERE transid = r.transid) AS  amount,
    (SELECT redeemedto FROM redeem_vouchers WHERE transid = r.transid LIMIT 0,1) AS 'Name' 
FROM 
    redeem_vouchers r 
WHERE 
    confirmedon 
BETWEEN ('2013-07-23 19:00:00') AND ('2013-07-24 19:00:00')

表只包含4909行，查询返回结果需要35秒。我添加了4个索引但没有找到速度增益。我知道我正在做一些危险但我是sql的新手。请帮帮我。

Answer 1

我会尝试使用GROUP BY的查询：

SELECT
  transid,
  COUNT(voucheramount)*16 AS Amount,
  redeemedto As `Name`
FROM
  redeem_vouchers
WHERE
  confirmedon BETWEEN '2013-07-23 19:00:00' AND '2013-07-24 19:00:00'
GROUP BY
  transid

redeemedto是一个非聚合列，其值将不确定，但在原始查询中也未确定，因为您使用LIMIT获取第一条记录，但您没有使用ORDER BY。

如果所有兑换的值都相同，那就没问题 - 或者您可以使用

MAX(redeemedto) As `Name`

使查询成为SQL标准。

请帮我优化包含子查询的mysql查询，只需几百条记录

1 个答案: