有一張很老的數(shù)據(jù)表，時(shí)間戳格式為varchar，字段如下：idbigintnamevarchar(200)create_timevarchar(200)//索引KEY`IDX_CREATED`(`create_time`),數(shù)據(jù)約500多萬(wàn)，現(xiàn)在引出發(fā)現(xiàn)的問(wèn)題，一條sql語(yǔ)句效率非常的低：selectid,namefromtwherecreate_time>1434115807296orderbycreate_timelimit1000;本機(jī)測(cè)試200s，執(zhí)行計(jì)劃：>explainselectid,namefromtwherecreate_time>1434115807296orderbycreate_timelimit1000;+----+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+|id|select_type|table|type|possible_keys|key|key_len|ref|rows|Extra|+----+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+|1|SIMPLE|User|index|IDX_CREATED|IDX_CREATED|63|NULL|1000|Usingwhere|+----+-------------+-------+-------+---------------+---------------+---------+------+------+-------------+1rowinset(0.00sec)如果去掉limit:selectid,namefromtwherecreate_time>1434115807296orderbycreate_time執(zhí)行時(shí)間5s，執(zhí)行計(jì)劃：>explainselectid,namefromtwherecreate_time>1434115807296orderbycreate_time+----+-------------+-------+------+---------------+------+---------+------+---------+-----------------------------+|id|select_type|table|type|possible_keys|key|key_len|ref|rows|Extra|+----+-------------+-------+------+---------------+------+---------+------+---------+-----------------------------+|1|SIMPLE|User|ALL|IDX_CREATED|NULL|NULL|NULL|4858500|Usingwhere;Usingfilesort|+----+-------------+-------+------+---------------+------+---------+------+---------+-----------------------------+1rowinset(0.00sec)一個(gè)index查詢(xún)竟然比ALL&filesort查詢(xún)慢這么多？請(qǐng)MySQL達(dá)人指教