# 快速定位查询存储倾斜的表
目前提供的倾斜查询接口有函数：[table_distribution(schemaname text, tablename text)](https://support.huaweicloud.com/distributed-devg-v2-gaussdb/gaussdb-12-0281.html#ZH-CN_TOPIC_0000002222888741__zh-cn_topic_0000002220898457_li740916348447)、[table_distribution()](https://support.huaweicloud.com/distributed-devg-v2-gaussdb/gaussdb-12-0281.html#ZH-CN_TOPIC_0000002222888741__zh-cn_topic_0000002220898457_li645551081820)以及视图[PGXC_GET_TABLE_SKEWNESS](https://support.huaweicloud.com/distributed-devg-v2-gaussdb/gaussdb-12-0831.html#ZH-CN_TOPIC_0000002187528996)，客户可以根据自身业务情况来选择使用。
#### 场景一：磁盘满后快速定位存储倾斜的表
首先，通过[pg_stat_get_last_data_changed_time(oid)](https://support.huaweicloud.com/distributed-devg-v2-gaussdb/gaussdb-12-0282.html#ZH-CN_TOPIC_0000002187368624__zh-cn_topic_0000002185458402_li1248175514307)函数查询出近期发生过数据变更的表，鉴于表的最后修改时间只在进行IUD操作的CN记录，要查询库内1天(间隔可在函数中调整)内被修改的所有表，可以使用如下封装函数：
```
CREATE OR REPLACE FUNCTION get_last_changed_table(OUT schemaname text, OUT relname text) 
RETURNS setof record 
AS $$ 
DECLARE 
row_data record; 
row_name record; 
query_str text; 
query_str_nodes text; 
BEGIN 
query_str_nodes := 'SELECT node_name FROM pgxc_node where node_type = ''C'''; 
FOR row_name IN EXECUTE(query_str_nodes) LOOP 
query_str := 'EXECUTE DIRECT ON (' || row_name.node_name || ') ''SELECT b.nspname,a.relname FROM pg_class a INNER JOIN pg_namespace b on a.relnamespace = b.oid where pg_stat_get_last_data_changed_time(a.oid) BETWEEN current_timestamp - 1 AND current_timestamp;'''; 
FOR row_data IN EXECUTE(query_str) LOOP 
schemaname = row_data.nspname; 
relname = row_data.relname; 
return next; 
END LOOP; 
END LOOP; 
return; 
END; $$ 
LANGUAGE 'plpgsql';
```
然后，通过[table_distribution(schemaname text, tablename text)](https://support.huaweicloud.com/distributed-devg-v2-gaussdb/gaussdb-12-0281.html#ZH-CN_TOPIC_0000002222888741__zh-cn_topic_0000002220898457_li740916348447)查询出表在各个DN占用的存储空间。
```
SELECT table_distribution(schemaname,relname) FROM get_last_changed_table();
```
#### 场景二：常规数据倾斜巡检
- 在库中表个数少于1W的场景，直接使用倾斜视图查询当前库内所有表的数据倾斜情况。
  ```
  SELECT * FROM pgxc_get_table_skewness ORDER BY totalsize DESC;
  ```
  

- 在库中表个数非常多（至少大于1W）的场景，因[PGXC_GET_TABLE_SKEWNESS](https://support.huaweicloud.com/distributed-devg-v2-gaussdb/gaussdb-12-0831.html#ZH-CN_TOPIC_0000002187528996)涉及全库查并计算非常全面的倾斜字段，所以可能会花费比较长的时间（小时级），建议参考[PGXC_GET_TABLE_SKEWNESS](https://support.huaweicloud.com/distributed-devg-v2-gaussdb/gaussdb-12-0831.html#ZH-CN_TOPIC_0000002187528996)视图定义，直接使用table_distribution()函数自定义输出，减少输出列进行计算优化，例如：
  ```
  SELECT schemaname,tablename,max(dnsize) AS maxsize, min(dnsize) AS minsize 
  FROM pg_catalog.pg_class c 
  INNER JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace 
  INNER JOIN pg_catalog.table_distribution() s ON s.schemaname = n.nspname AND s.tablename = c.relname 
  INNER JOIN pg_catalog.pgxc_class x ON c.oid = x.pcrelid AND x.pclocatortype = 'H' 
  GROUP BY schemaname,tablename;
  ```