1

12 個のテーブルの行を単純に「すべて結合」しようとしています。すべてのテーブルを組み合わせると、115 行になります。ただし、以下のクエリを実行すると、「explain」を使用すると次のようになり、「log-queries-not-using-indexes」に設定すると mysql-slow.log にエントリが記録されます。

Rows_sent: 115  Rows_examined: 1008

すべての行を結合 (つまり、「結合」) するだけでよいのに、mysql が 1008 行を調べる方法に少し戸惑っています。ヒントや指針をいただければ幸いです。

クエリは次のとおりです。

(SELECT id, var_lng_1, 0 as tbl_col FROM tbl1 )

UNION ALL 

(SELECT id, var_lng_1, 1 as tbl_col FROM tbl2 )

UNION ALL 

(SELECT id, var_lng_1, 2 as tbl_col FROM tbl3 )

UNION ALL 

(SELECT id, var_lng_1, 3 as tbl_col FROM tbl4 )

UNION ALL 

(SELECT id, var_lng_1, 4 as tbl_col FROM tbl5 )

UNION ALL 

(SELECT id, var_lng_1, 5 as tbl_col FROM tbl6 )

UNION ALL 

(SELECT id, var_lng_1, 6 as tbl_col FROM tbl7 )

UNION ALL 

(SELECT id, var_lng_1, 7 as tbl_col FROM tbl8 )

UNION ALL 

(SELECT id, var_lng_1, 8 as tbl_col FROM tbl9 )

UNION ALL 

(SELECT id, var_lng_1, 9 as tbl_col FROM tbl10 )

UNION ALL 

(SELECT id, var_lng_1, 10 as tbl_col FROM tbl11 )

UNION ALL 

(SELECT id, var_lng_1, 11 as tbl_col FROM tbl12 );

どんな入力でも大歓迎です

PS: (念のため) すべての ID は primary,tiny_int(3), auto_increment 列です。IDだけで同じクエリも試しました（つまり、「select id from ....」ですが、違いはありませんでした:(

完全な説明出力:

[id] => 1  
[select_type] => PRIMARY  
[table] => tbl1  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 7  
[Extra] =>   

[id] => 2  
[select_type] => UNION  
[table] => tbl2  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 18  
[Extra] =>   

[id] => 3  
[select_type] => UNION  
[table] => tbl3  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 8  
[Extra] =>   

[id] => 4  
[select_type] => UNION  
[table] => tbl4  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 10  
[Extra] =>   

[id] => 5  
[select_type] => UNION  
[table] => tbl5  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 11  
[Extra] =>   

[id] => 6  
[select_type] => UNION  
[table] => tbl6  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 14  
[Extra] =>   

[id] => 7  
[select_type] => UNION  
[table] => tbl7  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 10  
[Extra] =>   

[id] => 8  
[select_type] => UNION  
[table] => tbl8  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 6  
[Extra] =>   

[id] => 9  
[select_type] => UNION  
[table] => tbl9  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 3  
[Extra] =>   

[id] => 10  
[select_type] => UNION  
[table] => tbl10  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 2  
[Extra] =>   

[id] => 11  
[select_type] => UNION  
[table] => tbl11  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 6  
[Extra] =>   

[id] => 12  
[select_type] => UNION  
[table] => tbl12  
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] => 20  
[Extra] =>   


[id] =>   
[select_type] => UNION RESULT  
[table] => <union1,2,3,4,5,6,7,8,9,10,11,12>   
[type] => ALL  
[possible_keys] =>   
[key] =>   
[key_len] =>   
[ref] =>   
[rows] =>   
[Extra] =>
4

1 に答える 1

1

句がないWHEREため、すべての行を返します。UNION ALL( とは対照的にUNION) を使用すると、異なるセットでの重複のチェックが防止されます。ここでは、正しい結果を得るために、各テーブルの完全なテーブル スキャンが必要です。そして、あなたにもORDER BY節がありません。このクエリには、インデックスの恩恵を受ける可能性のあるものは何もありません。インデックスを使用するだけでは役に立ちません。

于 2010-12-10T05:09:41.447 に答える