假设我有一个巨大的表格,可以使用以下格式从一开始就保存大量日志:
==========================================================
| Name | Log | Date (type datetime) |
----------------------------------------------------------
| Bob | {Some:[sort,of,json]} | 1-May-2013 09:12:45 |
| Josh | {Another:[sort,of,json]} | 1-May-2013 09:13:45 |
| Fred | {Yada:[yada,yada,yada]} | 1-May-2013 09:14:45 |
| Josh | {Ahoy:[whee,whee,whee]} | 1-May-2013 09:15:45 |
| Lucy | {Ahem:[blagh,blgh,blgh]} | 1-May-2013 09:16:45 |
| Bob | {Chih:[aw,ew,ow]} | 2-May-2013 09:12:45 |
..........................................................
| Fred | {Cheh:[saw,sew,sow]} | 1-May-2014 09:12:45 |
| Bob | {Chah:[waw,wew,wow]} | 1-May-2014 09:15:45 |
==========================================================
现在,给定两个日期时间,我需要在这些日期时间之间为每个人抓一个日志(该时间内的任何日志都会这样做,但最好是在这两个日期时间内最早的日志)。
以下是我尝试过的查询,但仍需要很长时间
select *
from ( select Name,
Log,
rank() over (partition by Name order by Date asc) ranks
from Table ) alias
where ranks = 1
答案 0 :(得分:1)
您必须使用BETWEEN
SELECT *
FROM (
SELECT Name, Log,
rank() OVER (partition by Name ORDER by Date ASC) ranks
FROM Table
WHERE Date BETWEEN '2012-04-01 02:00:00′ AND '2012-04-20 02:00:00′
) alias
WHERE ranks = 1;
您可能要在用于加速查询执行的字段上创建INDEX
。
答案 1 :(得分:0)
试试这个
SELECT *
FROM [YourTable]
WHERE [YourDate] BETWEEN '08-15-2011 02:00:00' AND '08-16-2011 02:00:00'
答案 2 :(得分:0)
您应该使用BETWEEN条件但在子查询内部
select *
from ( select Name,
Log,
rank() over (partition by Name order by Date asc) ranks
from Table
WHERE Date BETWEEN @DateBegin AND @DateEnd
) alias
where ranks = 1
要更快地运行它,您必须在Name
和Date
字段
答案 3 :(得分:0)
根据您提供的信息,我给出了一个通用示例:
SELECT *
FROM [YourTable]
WHERE [Your Date] BETWEEN '1-May-2013 02:00:00' AND '2-May-2013 02:00:00'
ORDER BY [yourDate] DESC