首先，您的查询可以用一种更简单的方式表示为：
SELECT COUNT(*) AS totalCalls, HOUR(`end`) AS `Hour`
FROM callsDataTable c
INNER JOIN products p ON c.company = p.number
AND p.id IN (@_PRODUCTS)
AND YEAR(`end`) = @_YEAR AND MONTH(`end`) = @_MONTH
GROUP BY HOUR(`end`) AS `Hour`
ORDER BY `Hour` ASC
使用@NoDisplayName在他们的答案中提出的想法：
CREATE TABLE hours_table (hours INT);
INSERT INTO hours_table VALUES(0), (1), (2),
/* put the missing values here */ (23);
您可以加入包含小时数的表，以获取所需的结果：
SELECT COUNT(*) AS totalCalls, h.hours AS `Hour`
FROM callsDataTable c
INNER JOIN products p ON c.company = p.number
RIGHT JOIN hours_table h ON h.hours = HOUR(c.`end`)
AND p.id IN (@_PRODUCTS)
AND YEAR(`end`) = @_YEAR AND MONTH(`end`) = @_MONTH
GROUP BY h.hours
ORDER BY h.hours ASC
如果运行太慢（我确定运行速度很慢），则应该研究一种使用类似方法的方法，end BETWEEN '2015-01-01 00:00:00' AND '2015-01-31 23:59:59'而不是比较YEAR(end)和MONTH(end)。
可以这样完成：
SET @start = STR_TO_DATE(CONCAT(@_YEAR, '-', @_MONTH, '-01 00:00:00'), '%Y-%m-%d %H:%i:%s');
SET @end = DATE_SUB(DATE_ADD(@start, INTERVAL 1 MONTH), INTERVAL 1 SECOND);
SELECT ...
...
AND `end` BETWEEN @start AND @end
...
但是这种改变本身并没有帮助。它需要一个现场索引end来带来所需的速度改进：
ALTER TABLE callsDataTable ADD INDEX(end);
使用HOUR(c.end)的连接条件是另一个原因运行缓慢。
可以通过将表hours_table与第一个查询（的简化版本）生成的结果集结合在一起来进行改进：
SELECT IFNULL(totalCalls, 0) AS totalCalls, h.hours AS `Hour`
FROM hours_table h
LEFT JOIN (
SELECT COUNT(*) AS totalCalls, HOUR(`end`) as `Hour`
FROM callsDataTable c
INNER JOIN products p ON c.company = p.number
AND p.id IN (@_PRODUCTS)
AND YEAR(`end`) = @_YEAR AND MONTH(`end`) = @_MONTH
GROUP BY HOUR(`end`) AS `Hour`
) d ON h.hours = d.`Hour`
ORDER BY h.hours ASC
602392714
清零编程群