MySql，JOIN 和 Group By 查询正在使用临时和文件排序

Question

I have 2 tables as below:

CREATE TABLE `ox_campaigns` (                                                       
    `campaignid` mediumint(9) NOT NULL auto_increment,                                
    `campaignname` varchar(255) NOT NULL default '',                                  
    `clientid` mediumint(9) NOT NULL default '0',                                                                                                    
    `is_deleted` tinyint(1) NOT NULL default '0',                                        
    PRIMARY KEY  (`campaignid`),                                                      
    KEY `ox_campaigns_clientid` (`clientid`)                                          
) ENGINE=MyISAM DEFAULT CHARSET=utf8  

CREATE TABLE `ox_clients` (                                   
  `clientid` mediumint(9) NOT NULL auto_increment,            
  `agencyid` mediumint(9) NOT NULL default '0',               
  `clientname` varchar(255) NOT NULL default '',                                                                
  `is_deleted` tinyint(4) NOT NULL,                           
  PRIMARY KEY  (`clientid`),                                  
  UNIQUE KEY `ox_clients_account_id` (`account_id`),          
  KEY `ox_clients_agencyid` (`agencyid`)                      
) ENGINE=MyISAM DEFAULT CHARSET=utf8              

One client can have many campaigns linked to it.

I have a list of campaignids with me, what I want is a list of distinct clientids for those campaigns.

The query that I am using is:

SELECT clients.* 
    FROM clients  
    JOIN campaigns ON clients.clientid = campaigns.clientid
    WHERE campaigns.is_deleted=0 
        AND campaignid in (2325,2395)
        AND clients.is_deleted=0 
    GROUP BY clients.clientid

The EXPLAIN output that it gives is:

+----+-------------+-----------+-------+-------------------------------+---------+--------+-------------------------------------------------------------------------------+
| id | select_type | table     | type  | possible_keys                 | key     | key_len| ref                      | rows| Extra                                        |
+----+-------------+-----------+-------+-------------------------------+---------+--------+-------------------------------------------------------------------------------+
| 1  |    SIMPLE   | campaigns | range | PRIMARY,ox_campaigns_clientid | PRIMARY | 3      |             NULL         | 2   | Using where; Using temporary; Using filesort |
| 1  |    SIMPLE   | clients   | eq_ref| PRIMARY                       | PRIMARY | 3      | openx.campaigns.clientid | 1   | Using where

Why it is using temporary and filesorting for this query?

Answer 1

It is using filesort because of the group by . You can prevent this by using an exists clause for what you are doing:

SELECT c.* 
FROM clients c 
WHERE EXISTS (SELECT 1
              FROM campaigns ca
              WHERE ca.clientid = c.clientid AND
                    ca.is_deleted = 0 AND
                    ca.campaignid IN (2325,2395)
             )
      AND c.is_deleted = 0;

You have an index on campaigns_clientid(clientid) , so this should use the index. A better index would be campaigns_clientid(clientid, is_deleted, campaign_id) . This index "covers" the subquery. In other words, the engine will only use the index and not have to read in the data pages for that table.