Question

我甚至不知道我的头衔是否有意义，所以希望我的解释能够弥补我的标题清晰的湖泊。

我想创建一个查询，返回有关我的表的一般信息，行数，平均长度，数据长度，最小值（id）和最大值（id）等。

我知道如何查询INFORMATION_SCHEMA.PARTITIONS以获取'尺寸'信息。例如：

mysql> SELECT TABLE_NAME AS Name,
    ->     TABLE_ROWS as 'Row Count',
    ->     AVG_ROW_LENGTH AS 'Avg Row Len',
    ->     DATA_LENGTH,
    ->     ROUND(DATA_LENGTH / 1024 / 1024) AS 'DATA_LENGTH(M)'
    -> 
    -> FROM INFORMATION_SCHEMA.PARTITIONS
    -> 
    -> WHERE (TABLE_NAME    = 'event_log'
    ->        OR TABLE_NAME = 'event_log_policy_entries'
    ->        OR TABLE_NAME = 'event_log_policy_status');
+--------------------------+-----------+-------------+-------------+----------------+
| Name                     | Row Count | Avg Row Len | DATA_LENGTH | DATA_LENGTH(M) |
+--------------------------+-----------+-------------+-------------+----------------+
| event_log                |  10000089 |         182 |  1822425088 |           1738 |
| event_log_policy_entries |  10000137 |         171 |  1710227456 |           1631 |
| event_log_policy_status  |  10000094 |         244 |  2449473536 |           2336 |
+--------------------------+-----------+-------------+-------------+----------------+
3 rows in set (0.57 sec)

mysql>

此外，我知道如何在每个表的基础上使用MIN()和MAX()函数来获取其最小/最大ID值。（我的所有三个表都有一个名为id的列。）

mysql> SELECT MIN(id) AS 'Min', MAX(id) AS 'Max' from event_log;
+------+---------+
| Min  | Max     |
+------+---------+
|    0 | 9999999 |
+------+---------+
1 row in set (0.00 sec)

mysql>

但我不知道如何使这个 一个查询 。我的目标是拥有一个如下表所示的表：

+--------------------------+-----------+-------------+-------------+----------------+-----+-----+
| Name                     | Row Count | Avg Row Len | DATA_LENGTH | DATA_LENGTH(M) | Min | Max |
+--------------------------+-----------+-------------+-------------+----------------+-----+-----+
| event_log                |  10000089 |         182 |  1822425088 |           1738 |   # |   # |
| event_log_policy_entries |  10000137 |         171 |  1710227456 |           1631 |   # |   # |
| event_log_policy_status  |  10000094 |         244 |  2449473536 |           2336 |   # |   # |
+--------------------------+-----------+-------------+-------------+----------------+-----+-----+

我一直在阅读关于表JOIN和SUBQUERIES的内容，但我还没有弄清楚这一点。（我假设这可以通过单个查询完成）。

Answer 1

试试这个：

mysql> SELECT TABLE_NAME AS Name,
    ->     TABLE_ROWS as 'Row Count',
    ->     AVG_ROW_LENGTH AS 'Avg Row Len',
    ->     DATA_LENGTH,
    ->     ROUND(DATA_LENGTH / 1024 / 1024) AS 'DATA_LENGTH(M)',
    ->     MIN(id) AS 'Min', MAX(id) AS 'Max'
    -> 
    -> FROM INFORMATION_SCHEMA.PARTITIONS, event_log
    -> 
    -> WHERE (TABLE_NAME    = 'event_log'
    ->        OR TABLE_NAME = 'event_log_policy_entries'
    ->        OR TABLE_NAME = 'event_log_policy_status');

它将返回INFORMATION_SCHEMA.PARTITIONS的过滤结果，以及event_log的所有结果。

我遇到的一个问题是，event_log表中是否有一个与INFORMATION_SCHEMA.PARTITIONS表相关的字段，以便您可以将event_log的每一行绑定到INFORMATION_SCHEMA.PARTITIONS表中的一行？

Answer 2

不是堆栈溢出用户的同事向我推荐了一个示例，然后我将其推断为以下内容：

mysql> SELECT TABLE_NAME AS Name,
    ->     AVG_ROW_LENGTH AS 'Avg_Row_Size',
    ->     DATA_LENGTH AS 'Table_Size',
    ->     INDEX_LENGTH AS 'Index_Size',
    ->     elog.Min,
    ->     elog.Max
    -> 
    -> FROM INFORMATION_SCHEMA.PARTITIONS AS isp
    -> 
    -> LEFT JOIN
    ->   (SELECT 'event_log' AS 'Name',
    ->           MIN(id) AS 'Min',
    ->           MAX(id) AS 'Max'
    ->    FROM event_log
    ->    UNION
    ->    (SELECT 'event_log_policy_entries' AS 'Name',
    ->            MIN(id) AS 'Min',
    ->            MAX(id) AS 'Max'
    ->     FROM event_log_policy_entries)
    ->     UNION
    ->     (SELECT 'event_log_policy_status' AS 'Name',
    ->             MIN(id) AS 'Min',
    ->             MAX(id) AS 'Max'
    ->      FROM event_log_policy_status)) AS elog ON (elog.Name = isp.TABLE_NAME)
    -> 
    -> WHERE (isp.TABLE_NAME    = 'event_log'
    ->        OR isp.TABLE_NAME = 'event_log_policy_entries'
    ->        OR isp.TABLE_NAME = 'event_log_policy_status');
+--------------------------+--------------+------------+------------+------+---------+
| Name                     | Avg_Row_Size | Table_Size | Index_Size | Min  | Max     |
+--------------------------+--------------+------------+------------+------+---------+
| event_log                |          182 | 1822425088 |          0 |   11 | 9999999 |
| event_log_policy_entries |          170 | 1709178880 |  137019392 |   33 | 9999999 |
| event_log_policy_status  |          244 | 2448392192 |  137019392 |  132 | 9999999 |
+--------------------------+--------------+------------+------------+------+---------+
3 rows in set (0.56 sec)

mysql>

它似乎与原始查询一样快，没有最小值/最大值。我使用SELECT和UNION创建了一个包含名称和最小/最大值的子表，然后将其连接到INFORMATION_SCHEMA.PARTITION TABLE_NAME的结果。

有更好的方法吗？这很有效，我觉得这很好。但是有更好的方法吗？

如何使用一个表中的行值作为一个查询中另一个表的名称？

2 个答案: