How to use MySQL to Limit secondary results

I've been thinking on this one and it's stumped me.  

I need to query a user table (dhcpclients) for all users (31 records) and include 25 rows for each user from another table (logs). So the end result is 775 rows- 25 for each 31. Make sense?

Attached is the code I'm working with.  

The query that makes up the "join" in there works fine but obviously stops at the limit- which is this case would be one user record with 25 of the log table records. the problem is if I move the limit up it's not going to move to the next user until it's exhausted all the matches in the log table. I need it to quit after 25 and move to the next user.

Make sense?


SELECT dhcp.machinename, log1.* FROM dhcpclients dhcp
JOIN (
SELECT DATE(logs.date) AS 'Date', SUBSTRING_INDEX(SUBSTRING_INDEX(logs.request,'://',-1),'/',1) AS 'Site',
COUNT(logs.request) AS 'Hits',
FORMAT(SUM(logs.elapsed)/1000,2) AS 'Xfr Time(sec)', FORMAT(SUM(logs.bytes)/1024,2) AS 'KB Txfrd',
FORMAT((((SUM(logs.bytes)/1024)/1024)*8)/(SUM(logs.elapsed)/1000),2) AS 'Mbits/s', logs.ip
FROM logs 
WHERE logs.request NOT LIKE '[%' AND SUBSTRING_INDEX(SUBSTRING_INDEX(logs.request,'://',-1),'/',1) NOT LIKE '%knothe.com%'
AND DATE(logs.date) = DATE(NOW())
GROUP BY DATE(logs.date),logs.ip, SUBSTRING_INDEX(SUBSTRING_INDEX(logs.request,'://',-1),'/',1)
ORDER BY DATE(logs.date) DESC, COUNT(logs.request)DESC, SUM(bytes)/1024 DESC
LIMIT 25 ) 

log1 ON log1.ip = dhcp.ipaddress 

ORDER BY dhcp.machinename ASC

Open in new window

LVL 1
digitalwavIT Infrastructure ManagerAsked:
Who is Participating?
 
APNFSSCConnect With a Mentor Commented:
digitalwav I think what you are after is a store procedure, so that you can loop through the subnets.

http://www.mysqltutorial.org/mysql-stored-procedure-tutorial.aspx

0
 
APNFSSCCommented:
Does your logs table have a primary auto incrementing key?

In this thread a method is described but it looks to me that it requires the key to work.

http://forums.mysql.com/read.php?20,135077,135077#msg-135077

First post then result

http://forums.mysql.com/read.php?20,135077,135107#msg-135107

0
 
raulggonzalezCommented:
Hi,

This works for MSSQL, but maybe you can make it work in MySQL ...

It's just an example with 2 tables, parent and child 1 to many.


Hope it helps
SELECT *  FROM 
	(SELECT RANK() OVER 
			(PARTITION BY Parent.ID ORDER BY Child.ID ASC) AS [RANK], Parent.ID, Child.ID as childID
		FROM XXXXX AS Parent
			INNER JOIN YYYYY AS Child
				ON Child.parentID = Parent.ID
	) AS TB
WHERE TB.[RANK] <= 25

Open in new window

0
Cloud Class® Course: Ruby Fundamentals

This course will introduce you to Ruby, as well as teach you about classes, methods, variables, data structures, loops, enumerable methods, and finishing touches.

 
digitalwavIT Infrastructure ManagerAuthor Commented:
Yes it has a auto increment key. I'll check out the thread you're referring to.
0
 
digitalwavIT Infrastructure ManagerAuthor Commented:
Ok, I tried the having count(xx.xx) idea and the results don't make any sense.  The rank idea doesn't work unless the rank can reset after each change in column 'machine name'

This simply may not be possible in pure sql. I may have to do it in php.
0
 
digitalwavIT Infrastructure ManagerAuthor Commented:
Let me try to explain this another way. I want the top 25 sites by machine_name. In this case the query I have gives ALL sites by machine_name for the day. If I just add limit to the statement it limits the overall results to 25, not the result by machine_name.
0
 
raulggonzalezCommented:
the idea of RANK is exactly what you want, I tried in my MS SQL database and it will return exactly what you need.

SELECT RANK() OVER
                  (PARTITION BY Parent.ID ORDER BY Child.ID ASC)

Gives a rank for every parentID so you will have
 
Parent   Child
1             1
1             2
1             3
2             1
2             2

and so on...

however I don't know how you can 'translate' this into MySql ...

Cheers
0
 
raulggonzalezCommented:
Sorry, the name of the columns is not right, It'll be like this

Parent   [RANK]
1             1
1             2
1             3
2             1
2             2
0
 
digitalwavIT Infrastructure ManagerAuthor Commented:
I'll see what I can come up with. MySQL can do something similar to Rank. I'll show my work when done.
0
 
digitalwavIT Infrastructure ManagerAuthor Commented:
The rank idea doesn't work. The only rank method for mysql is basically a counter 'column' which results in each record getting a consecutively higher number.  There is no way to reset that counter that I can see when then parent column changes to a new value.
0
 
digitalwavIT Infrastructure ManagerAuthor Commented:
yup, APNFSSC, that's what I'm going to have to do. Thanks for the pointer. I was hoping to avoid procedures.
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.