mysql 数据库统计出现误差, 在spring的调度器和程序逻辑都正常的情况下, 可以使用如下方法

1,查看调度器里面的配置, 查到数据的根位置,可以直接写储存过程来对统计表进行重新统计,, 如下, 我只贴一个例子--->

统计表能直接用联合查询可以把字段列出的:::------.>如下

DELIMITER $$


USE `bbbb`$$


DROP PROCEDURE IF EXISTS `A0_history_user_gdata_statistics_total_count_by_age_range_proc`$$


CREATE DEFINER=`root`@`%` PROCEDURE `A0_history_user_gdata_statistics_total_count_by_age_range_proc`()
BEGIN
  DECLARE vId VARCHAR (50) ;
  DECLARE vOrgId INTEGER (50) ;
  DECLARE vOrgPath VARCHAR (40) ;
  DECLARE vTime DATETIME ;
  DECLARE vQuantity INT ;
  DECLARE vAgeRange INTEGER(2) ;
  -- 遍历数据结束标志
  DECLARE done INT DEFAULT 0;
  DECLARE iCount INT ;
  DECLARE cur CURSOR FOR 
  SELECT 
 u.`organizationId`,
 o.`organizationPath`,
 MAX(u.`CREATE_TIME`) creatTime,
 CASE
WHEN (DATE_FORMAT(NOW(), '%Y')-LEFT(i.`BIRTHDAY`,4)) < 18 THEN 1
WHEN ((DATE_FORMAT(NOW(), '%Y')-LEFT(i.`BIRTHDAY`,4)) BETWEEN 18 AND 29) THEN 2
WHEN ((DATE_FORMAT(NOW(), '%Y')-LEFT(i.`BIRTHDAY`,4)) BETWEEN 30 AND 39) THEN 3
WHEN ((DATE_FORMAT(NOW(), '%Y')-LEFT(i.`BIRTHDAY`,4)) BETWEEN 40 AND 49) THEN 4
WHEN ((DATE_FORMAT(NOW(), '%Y')-LEFT(i.`BIRTHDAY`,4)) BETWEEN 50 AND 59) THEN 5
WHEN ((DATE_FORMAT(NOW(), '%Y')-LEFT(i.`BIRTHDAY`,4)) BETWEEN 60 AND 69) THEN 6
WHEN ((DATE_FORMAT(NOW(), '%Y')-LEFT(i.`BIRTHDAY`,4)) BETWEEN 70 AND 150) THEN 7
ELSE 
CASE
WHEN (DATE_FORMAT(NOW(), '%Y')-SUBSTRING(u.`IDCARD`,7,4)) < 18 THEN 1
WHEN ((DATE_FORMAT(NOW(), '%Y')-SUBSTRING(u.`IDCARD`,7,4)) BETWEEN 18 AND 29) THEN 2
WHEN ((DATE_FORMAT(NOW(), '%Y')-SUBSTRING(u.`IDCARD`,7,4)) BETWEEN 30 AND 39) THEN 3
WHEN ((DATE_FORMAT(NOW(), '%Y')-SUBSTRING(u.`IDCARD`,7,4)) BETWEEN 40 AND 49) THEN 4
WHEN ((DATE_FORMAT(NOW(), '%Y')-SUBSTRING(u.`IDCARD`,7,4)) BETWEEN 50 AND 59) THEN 5
WHEN ((DATE_FORMAT(NOW(), '%Y')-SUBSTRING(u.`IDCARD`,7,4)) BETWEEN 60 AND 69) THEN 6
WHEN ((DATE_FORMAT(NOW(), '%Y')-SUBSTRING(u.`IDCARD`,7,4)) BETWEEN 70 AND 150) THEN 7
ELSE 
   CASE
WHEN  i.`AGE` < 18 THEN 1
WHEN ( i.`AGE` BETWEEN 18 AND 29) THEN 2
WHEN ( i.`AGE` BETWEEN 30 AND 39) THEN 3
WHEN ( i.`AGE` BETWEEN 40 AND 49) THEN 4
WHEN ( i.`AGE` BETWEEN 50 AND 59) THEN 5
WHEN ( i.`AGE` BETWEEN 60 AND 69) THEN 6
WHEN ( i.`AGE` BETWEEN 70 AND 150) THEN 7
ELSE 3
   END
  END
 END ageRange,
 COUNT(1) AS quantity       
FROM
 `t_user_gdata` u,`t_user_organization` o,`t_user_info` i
WHERE u.`organizationId`=o.`organizationId`
AND u.`ID`=i.`USERID`
-- AND DATE_FORMAT(u.`CREATE_TIME`,'%Y-%m-%d') = DATE_ADD(DATE_FORMAT(NOW(),'%Y-%m-%d'),INTERVAL -1 DAY)
GROUP BY u.`organizationId`,o.`organizationPath`,ageRange;
  DECLARE CONTINUE HANDLER FOR SQLSTATE '02000' SET done = 1 ;
  /*删除历史数据*/
  DELETE 
  FROM
    t_user_gdata_statistics_total_count_by_age_range ;
  #set iCount = 0;
  /*对历史数据进行重新统计*/
  OPEN cur ;
  REPEAT
    FETCH cur INTO vOrgId,
    vOrgPath,
    vTime,
    vAgeRange,
    vQuantity;
    IF done = 0 THEN
    SET vId = REPLACE(UUID(), '-', '') ;
    INSERT INTO t_user_gdata_statistics_total_count_by_age_range (
      id,
      ageRange,
      quantity,
      organizationId,
      organizationPath
    ) VALUE (
      vId,
      vAgeRange,
      vQuantity,
      vOrgId,
      vOrgPath
    ) ;
    #set iCount = iCount + 1;
    #select CONCAT('循环次数:', iCount) ;
    END IF;
    UNTIL done 
  END REPEAT ;
  CLOSE cur ;
END$$


DELIMITER ;

如果有两张表,可以写两个存储过程 在删除统计表数据的时候任选一张::如下----->

DELIMITER $$


USE `bbbb`$$


DROP PROCEDURE IF EXISTS `A0_history_user_gdata_statistics_measurement_by_jiankangting`$$


CREATE DEFINER=`root`@`%` PROCEDURE `A0_history_user_gdata_statistics_measurement_by_jiankangting`()
BEGIN
  DECLARE vId VARCHAR (50) ;
  DECLARE vTimes VARCHAR (10) ;
  DECLARE vNormalQuantity INT(11) ;
  DECLARE vBussinessTypeId INT(11);
  DECLARE vOrganizationId INT (11);
  DECLARE vOrganizationPath TEXT ;
  -- 遍历数据结束标志
  DECLARE done INT DEFAULT 0;
  DECLARE iCount INT ;
  DECLARE cur CURSOR FOR 
-- 查询健康亭每个测量项的总正常次数
SELECT 
DATE_FORMAT (t.testDate,'%Y-%m-%d') times,
COUNT(*) normalQuantity, 
t.`bussinessTypeId`  bussinessTypeId,
t.`organizationId`  organizationId, 
t.organizationPath  organizationPath
FROM 
`t_user_measurement_ref` t
WHERE 
t.`userGroupId` 
IN (3, 4, 16, 19, 25, 36, 40, 48)

AND t.`source` = 17 GROUP BY t.`bussinessTypeId`, t.`organizationId`, t.`organizationPath`; 
 
  DECLARE CONTINUE HANDLER FOR SQLSTATE '02000' SET done = 1 ;
  /*删除历史数据*/
  DELETE 
  FROM
    t_user_measurement_total_by_jiankangting ;
  #set iCount = 0;
  /*对历史数据进行重新统计*/
  OPEN cur ;  
  REPEAT
    FETCH cur INTO 
    vTimes,
    vNormalQuantity,
    vBussinessTypeId,
    vOrganizationId,
    vOrganizationPath;
    
    IF done = 0 THEN
    SET vId = REPLACE(UUID(), '-', '') ;
    INSERT INTO t_user_measurement_total_by_jiankangting (
      id,
   times, 
  normalQuantity,
  bussinessTypeId,
  organizationId,
  organizationPath
    ) VALUE (
      vId,
      vTimes,
    vNormalQuantity ,
    vBussinessTypeId,
    vOrganizationId,
    vOrganizationPath
    ) ;
    #set iCount = iCount + 1;
    #select CONCAT('循环次数:', iCount) ;
    END IF;
    UNTIL done 
  END REPEAT ;
  CLOSE cur ;
END$$


DELIMITER ;

2.

DELIMITER $$


USE `bbbb`$$


DROP PROCEDURE IF EXISTS `A0_history_user_gdata_statistics_measuringabnormal_by_beikangyi`$$


CREATE DEFINER=`root`@`%` PROCEDURE `A0_history_user_gdata_statistics_measuringabnormal_by_beikangyi`()
BEGIN
  DECLARE vId VARCHAR (50) ;
  DECLARE vTimes VARCHAR (10) ;
  DECLARE vNormalQuantity INT(11) ;
  DECLARE vBussinessTypeId INT(11);
  DECLARE vOrganizationId INT (11);
  DECLARE vOrganizationPath TEXT ;
  -- 遍历数据结束标志
  DECLARE done INT DEFAULT 0;
  DECLARE iCount INT ;
  DECLARE cur CURSOR FOR 
-- 先查询到前一天倍康仪每个测量项的总异常次数 
SELECT
DATE_FORMAT (t.testDate,'%Y-%m-%d') times,
COUNT(*) abnormalQuantity, 
t.`bussinessTypeId`,
t.`organizationId`, 
t.`organizationPath`
FROM 
`t_user_measuringabnormal_ref` t
WHERE 
-- DATE_FORMAT(t.`testDate`, '%Y-%m-%d') = DATE_FORMAT(DATE_ADD(NOW(), INTERVAL - 1 DAY),'%Y-%m-%d') AND 
t.`source` = 9 
GROUP BY t.`bussinessTypeId`, t.`organizationId`, t.`organizationPath`; 
 
DECLARE CONTINUE HANDLER FOR SQLSTATE '02000' SET done = 1 ;
  /*之前数据已经删除过历史数据*/
 -- DELETE 
 -- FROM
  --  t_user_measurement_total_by_beikangyi ;
  #set iCount = 0;
  /*对历史数据进行重新统计*/
  OPEN cur ;  
  REPEAT
    FETCH cur INTO 
    vTimes,
    vNormalQuantity,
    vBussinessTypeId,
    vOrganizationId,
    vOrganizationPath;
    
    IF done = 0 THEN
    SET vId = REPLACE(UUID(), '-', '') ;
    INSERT INTO t_user_measurement_total_by_beikangyi (
      id,
   times, 
  abnormalQuantity,
  bussinessTypeId,
  organizationId,
  organizationPath
    ) VALUE (
      vId,
      vTimes,
    vNormalQuantity ,
    vBussinessTypeId,
    vOrganizationId,
    vOrganizationPath
    ) ;
    #set iCount = iCount + 1;
    #select CONCAT('循环次数:', iCount) ;
    END IF;
    UNTIL done 
  END REPEAT ;
  CLOSE cur ;
END$$


DELIMITER ;

涉及两张表的话在系统执行的时候必须执行先删除的操作.

经验之谈..