How to optimize some INNER JOIN Mysql query

Question

I am having some issue on my website that has every day more traffic and I was wondering how to optimize this type of query to MYSQL DB.

Query_time: 37 
Lock_time: 2 
Rows_sent: 10 
Rows_examined: 354287
SET timestamp=1323180396;

Query:

Results for explain extended SELECT a.*, UNIX_TIMESTAMP(a.createdon) AS timestamp, ct.cityname, ct.SeoCityName, scat.subcatname, scat.SeoSubcatName, cat.catid, cat.SeoCatName, cat.catname, 
          COUNT(*) AS piccount, p.picfile, UNIX_TIMESTAMP(feat.featuredtill) AS featuredtill 
            FROM prefix_ads a
                 INNER JOIN prefix_es_cities ct ON a.cityid = ct.cityid
         INNER JOIN prefix_es_subcats scat ON a.subcatid = scat.subcatid
         INNER JOIN prefix_es_cats cat ON scat.catid = cat.catid
        LEFT OUTER JOIN prefix_adxfields axf ON a.adid = axf.adid
                LEFT OUTER JOIN prefix_adpics p ON a.adid = p.adid AND p.isevent = '0'        
                LEFT OUTER JOIN prefix_es_featured feat ON a.adid = feat.adid AND feat.adtype = 'A'        
            WHERE scat.catid = 2 
                AND a.enabled = '1' AND a.verified = '1' AND a.expireson >= NOW()                   
            GROUP BY a.adid
            ORDER BY a.createdon DESC
            LIMIT 0, 18:

There we go with the EXPLAIN (I ma usinng PDO for my DB connection)

Array
(
    [id] => 1
    [0] => 1
    [select_type] => SIMPLE
    [1] => SIMPLE
    [table] => cat
    [2] => cat
    [type] => const
    [3] => const
    [possible_keys] => PRIMARY
    [4] => PRIMARY
    [key] => PRIMARY
    [5] => PRIMARY
    [key_len] => 2
    [6] => 2
    [ref] => const
    [7] => const
    [rows] => 1
    [8] => 1
    [filtered] => 100.00
    [9] => 100.00
    [Extra] => Using temporary; Using filesort
    [10] => Using temporary; Using filesort
)
Array
(
    [id] => 1
    [0] => 1
    [select_type] => SIMPLE
    [1] => SIMPLE
    [table] => scat
    [2] => scat
    [type] => ref
    [3] => ref
    [possible_keys] => PRIMARY,catid
    [4] => PRIMARY,catid
    [key] => catid
    [5] => catid
    [key_len] => 2
    [6] => 2
    [ref] => const
    [7] => const
    [rows] => 12
    [8] => 12
    [filtered] => 100.00
    [9] => 100.00
    [Extra] => 
    [10] => 
)
Array
(
    [id] => 1
    [0] => 1
    [select_type] => SIMPLE
    [1] => SIMPLE
    [table] => a
    [2] => a
    [type] => ref
    [3] => ref
    [possible_keys] => subcatid,cityid,verified,enabled,expireson
    [4] => subcatid,cityid,verified,enabled,expireson
    [key] => subcatid
    [5] => subcatid
    [key_len] => 2
    [6] => 2
    [ref] => some_DB.scat.subcatid
    [7] => some_DB.scat.subcatid
    [rows] => 354287
    [8] => 354287
    [filtered] => 100.00
    [9] => 100.00
    [Extra] => Using where
    [10] => Using where
)
Array
(
    [id] => 1
    [0] => 1
    [select_type] => SIMPLE
    [1] => SIMPLE
    [table] => ct
    [2] => ct
    [type] => eq_ref
    [3] => eq_ref
    [possible_keys] => PRIMARY
    [4] => PRIMARY
    [key] => PRIMARY
    [5] => PRIMARY
    [key_len] => 2
    [6] => 2
    [ref] => some_DB.a.cityid
    [7] => some_DB.a.cityid
    [rows] => 1
    [8] => 1
    [filtered] => 100.00
    [9] => 100.00
    [Extra] => 
    [10] => 
)
Array
(
    [id] => 1
    [0] => 1
    [select_type] => SIMPLE
    [1] => SIMPLE
    [table] => axf
    [2] => axf
    [type] => ref
    [3] => ref
    [possible_keys] => adid
    [4] => adid
    [key] => adid
    [5] => adid
    [key_len] => 4
    [6] => 4
    [ref] => some_DB.a.adid
    [7] => some_DB.a.adid
    [rows] => 1
    [8] => 1
    [filtered] => 100.00
    [9] => 100.00
    [Extra] => Using index
    [10] => Using index
)
Array
(
    [id] => 1
    [0] => 1
    [select_type] => SIMPLE
    [1] => SIMPLE
    [table] => p
    [2] => p
    [type] => ref
    [3] => ref
    [possible_keys] => adid
    [4] => adid
    [key] => adid
    [5] => adid
    [key_len] => 5
    [6] => 5
    [ref] => some_DB.a.adid,const
    [7] => some_DB.a.adid,const
    [rows] => 3
    [8] => 3
    [filtered] => 100.00
    [9] => 100.00
    [Extra] => 
    [10] => 
)
Array
(
    [id] => 1
    [0] => 1
    [select_type] => SIMPLE
    [1] => SIMPLE
    [table] => feat
    [2] => feat
    [type] => eq_ref
    [3] => eq_ref
    [possible_keys] => adid
    [4] => adid
    [key] => adid
    [5] => adid
    [key_len] => 7
    [6] => 7
    [ref] => some_DB.a.adid,const
    [7] => some_DB.a.adid,const
    [rows] => 1
    [8] => 1
    [filtered] => 100.00
    [9] => 100.00
    [Extra] => 
    [10] => 
)

More EXPLAIN from the same query

id  select_type     table   type    possible_keys   key     key_len     ref     rows    Extra
1   SIMPLE  cat     const   PRIMARY     PRIMARY     2   const   1   Using temporary; Using filesort
1   SIMPLE  scat    ref     PRIMARY,catid   catid   2   const   12  
1   SIMPLE  a   ref     subcatid,cityid,verified,enabled,expireson  subcatid    2   some_DB.scat.subcatid   354287  Using where
1   SIMPLE  ct  eq_ref  PRIMARY     PRIMARY     2   some_DB.a.cityid    1   
1   SIMPLE  axf     ref     adid    adid    4   some_DB.a.adid  1   Using index
1   SIMPLE  p   ref     adid    adid    5   some_DB.a.adid,const    3   
1   SIMPLE  feat    eq_ref  adid    adid    7   some_DB.a.adid,const    1 

Thanks

Answer 1

Try this query

SELECT 
    a.*, 
    UNIX_TIMESTAMP(a.createdon) AS timestamp, 
    ct.cityname, 
    ct.SeoCityName, 
    scat.subcatname, 
    scat.SeoSubcatName, 
    cat.catid, 
    cat.SeoCatName, 
    cat.catname, 
    COUNT(*) AS piccount, 
    p.picfile, 
    UNIX_TIMESTAMP(feat.featuredtill) AS featuredtill 
FROM 
    prefix_ads a
INNER JOIN 
    prefix_es_cities ct 
ON 
    a.enabled = '1' AND 
    a.verified = '1' AND 
    a.cityid = ct.cityid  AND 
    a.expireson >= NOW()                   
INNER JOIN 
    prefix_es_subcats scat 
ON 
    scat.catid = 2 AND a.subcatid = scat.subcatid
INNER JOIN 
    prefix_es_cats cat 
ON 
    scat.catid = cat.catid
LEFT OUTER JOIN 
    prefix_adxfields axf 
ON 
    a.adid = axf.adid
LEFT OUTER JOIN 
    prefix_adpics p 
ON 
    p.isevent = '0' AND a.adid = p.adid         
LEFT OUTER JOIN 
    prefix_es_featured feat 
ON 
    feat.adtype = 'A' AND a.adid = feat.adid         
GROUP BY 
    a.adid
ORDER BY 
    a.createdon DESC
LIMIT 0, 18;

Also create necessary indexes on the columns of the tables that are used in the joins and order by like prefix_ads.createdon , prefix_ads.adid ....

Hope this helps....