Sergy, you never cese to amaze me.... The both work perfect, both run at the same speed... ;)

Thanks again.


>Something like

SELECT CAST(dt.date1 + dt.min1 AS datetime) as Begins,
>	DATEADD( mi,10, CAST(dt.date1 + dt.min1 AS datetime)) AS Ends,
>	COUNT(*) as total FROM
>	 (SELECT proc_dt,
>	RIGHT(CONVERT( CHAR(8), proc_dt,108),5) as Time1,
>	CONVERT( CHAR(14), proc_dt,120) date1,
>	CASE
>	WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '50:00' THEN '50:00'
>	WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '40:00' THEN '40:00'
>	WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '30:00' THEN '30:00'
>	WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '20:00' THEN '20:00'
>	WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '10:00' THEN '10:00'
>	WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '00:00' THEN '00:00'
>	ELSE '??:??' END min1
>FROM orders
>WHERE proc_dt IS NOT NULL) dt
>GROUP BY CAST(dt.date1 + dt.min1 AS datetime), DATEADD( mi,10,
>	CAST(dt.date1 + dt.min1 AS datetime))
>ORDER BY 1
>
>-- or combaining things together
>SELECT dt.IntervalBegins,
>	DATEADD( mi,10, IntervalBegins) AS IntervalEnds,
>	COUNT(*) as total FROM
>	 (SELECT
>	   CAST( CONVERT( CHAR(14), proc_dt,120) +
>		CASE
>		WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '50:00' THEN '50:00'
>		WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '40:00' THEN '40:00'
>		WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '30:00' THEN '30:00'
>		WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '20:00' THEN '20:00'
>		WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '10:00' THEN '10:00'
>		WHEN RIGHT(CONVERT( CHAR(8), proc_dt,108),5) > '00:00' THEN '00:00'
>		ELSE '??:??' END AS datetime ) AS IntervalBegins
>FROM orders
>WHERE proc_dt IS NOT NULL) dt
>GROUP BY IntervalBegins, DATEADD( mi,10, IntervalBegins)
>ORDER BY 1

You convert datime field proc_dt into beginnig of the interval and than group by it.
>
>
>>Im havnig problems getting this query to work right. What I have is a table of logins to a website. In the table containes the dattime of the loging along with other info. I need a query containing the number of logins within every 10 minute time interval.
>>
>>For example
>>
>>1/1/02 12:00 pm to 1/1/02 12:10 pm - 10 logins
>>1/1/02 12:10 pm to 1/1/02 12:20 pm - 15 logins
>>1/1/02 12:20 pm to 1/1/02 12:30 pm - 6 logins
>>...
>>
>>I hope this makes sence. Any ideas?
>>
>>Eric Stephani