我有一个包含以下字段的表
Time | Status
2016-01-21 16:11:00 | Connected
2016-01-21 16:51:00 | Disconnected
2016-01-21 17:01:00 | Connected
2016-01-21 17:10:00 | Status X
2016-01-21 17:20:00 | Status Y
2016-01-21 17:25:00 | Disconnected
2016-01-21 17:30:00 | Connected
2016-01-21 17:32:00 | Disconnected
我希望输出像这样
Disconnected Connected
2016-01-21 16:51:00 | 2016-01-21 17:01:00
2016-01-21 17:25:00 | 2016-01-21 17:30:00
2016-01-21 17:32:00 |
等等。也就是说,我想显示服务器断开连接时间的数据。
我写过像这样的查询
select B.Time, A.Time
from ( select Time
from table
where Status = 'Connected') as A, table B
where B.Status = 'Disconnected'
and B.Time < A.Time;
我的输出是
Disconnected Connected
2016-01-21 16:51:00 2016-01-21 17:01:00
2016-01-21 16:51:00 2016-01-21 17:30:00
也就是说,对于每个断开连接的事件,我得到的所有连接事件都大于它。但是,我希望每行只有第一个连接事件大于disconnect事件。我该怎么办?
答案 0 :(得分:3)
一个好的方法是创建一个包含所有时间逻辑的更复杂的ON
子句。 (http://sqlfiddle.com/#!9/81f61/5/0)
select dis.Time as disconnected, con.Time as connected
from ev dis
join ev con ON con.Status = 'connected'
AND dis.Status = 'disconnected'
AND con.Time = (select MIN(Time)
from ev
where Status = 'connected'
and Time > dis.Time)
这会限制每个连接的行仅考虑晚于每个断开连接时间的第一个MIN()
连接时间。
如果您有大量这些事件记录,请考虑在(Status, Time)
上放置索引以优化此查询。
答案 1 :(得分:1)
您可以通过断开连接时间对输出进行分组,并使用min来获取最早的输出:
SELECT B.Time,
min(A.Time)
FROM
(SELECT TIME
FROM TABLE
WHERE Status = 'Connected') AS A,
TABLE B
WHERE B.Status = 'Disconnected'
AND B.Time < A.Time
GROUP BY 1
;
或没有子查询:
SELECT tab1.`time`,
min(tab2.`time`)
FROM tab1 AS tab1
LEFT JOIN tab1 AS tab2 ON tab2.time > tab1.time
WHERE tab1.Status = 'Disconnected'
GROUP BY 1
;
答案 2 :(得分:0)
之前我使用过此查询,似乎运行良好
SELECT B.Time, A.Time
FROM table
WHERE NOT EXISTS (
SELECT time
WHERE b.Time = a.Time
AND b.Disconnected > a.Connected
)