GreenPlum 大数据平台--分区表(二)

01,创建分区表

[PARTITION BY partition_type (column)
    [SUBPARTITION BY partition_type (column)]
        [SUBPARTITION TEMPLATE (template_spec)]
    [...]
    (partition_spec)
        |[SUBPARTITION BY partition_type(column)]
         [...]
    (partition_spec)
        [(subpartition_spec
            [(...)]
        )]
]
and partition_element is:

DEFAULT PARTITION name
| [PARTITION name] VALUES (list_value[,...])
| [PARTITION name]
    START ([datatype] 'start_value') [INCLUSIVE|EXCLUSIVE]
    [ END ([datatype] 'end_value') [INCLUSIVE|EXCLUSIVE]
    [ EVERY ([datatype] [number|INTERVAL] 'interval_value')]
| [PARTITION name]
    END ([DATATYPE] 'end_value') [INCLUSIVE|EXCLUSIVE]
    [ EVERY ([datatype] [number|INTERVAL] 'interval_value')]
    [ with (partition_storage_parameter=value [,...])]
    [ TABLESPACE tablespace]

Tips
通过实验得到：向主表插入数据时，数据会被自动存放至相应的分区表。
也可以直接向分区子表插入符合条件的数据，
当向分区子表插入不符合条件的额数据时，会提示：

[Err] ERROR: Trying to insert row into wrong partition (seg1 hadoop3:40000 pid=6679)
DETAIL: Expected partition: test_partition_range_1_prt_p20111231, provided partition: test_partition_range_1_prt_p20111230

02,按时间分区

create table public.test_partition_range(
  id numeric,
  name character varying(32),
  dw_end_date date,
) distributed by(id)
PARTITION BY range(dw_end_date)
(
    PARTITION p20111230 START ('2011-12-30'::date) END ('2011-12-31'::date),
    --...
);

03,使用Every分区

create table test.test_partition_every(
  id numeric,
  name character varying(32),
  dw_end_date date
) distributed by(id)
partition by range(dw_end_Date)
(
  partition p201112 start('2011-12-1'::date) end ('2011-12-31'::date)
  every ('1 days'::interval)
);

04,使用list分区

create table test.test_partition_list(
  member_id numeric,
  city character varying(32)
)distributed by (member_id)
partition by list(city)
(
  partition guangzhou values('guangzhou'),
  partition hangzhou values('hangzhou'),
  default partition other_city
);

05,修改分区表

ALTER DEFAULT PARTITION
DROP DEFAULT PARTITION [IF EXISTS]
DROP PARTITION [IF EXISTS] {
  partition_name
  | FOR (RANK(number))
  | FOR (value)
}
[CASCADE]
TRUNCATE DEFAULT PARTITION
TRUNCATE PARTITION {
  partition_name
  | FOR (RANK(number))
  | FOR (value)
}
RENAME DEFAULT PARTITION TO new_partition_name
RENAME PARTITION {
  partition_name
  | FOR (RANK(number))
  | FOR (value)
}
TO new_partition_name
ADD DEFAULT PARTITION NAME [(subpartition_spec)]
ADD PARTITION [name] partition_element
  [(subpartition_spec)]
EXCHANGE PARTITION {
  partition_name
  | FOR (RANK(number))
  | FOR (value)
} WITH TABLE TABLE_NAME
  [WITH|WITHOUT VALIDATION]
EXCHANGE EFAULT PARTITION WITH TABLE TABLE_NAME
  [WITH|WITHOUT VALIDATION]
SET SUBPARTITION TEMPLATE (subpartition_spec)
SPLIT DEFAULT PARTITION {
    AT (list_value)
  |START([datatype] range_value) [INCLUSIVE|EXCLUSIVE]
   END ([datatype]) range_value) [INCLUSIVE|EXCLUSIVE]
}
[INTO (PARTITION new_partition_name,
       PARTITION default_partition_name)]
SPLIT DPARTITION {
  partition_name
  | FOR (RANK(number))
  | FOR (value)
} AT(value)
[INTO (PARTITION partition_name, PARTITION partition_name]

　　06,新增分区

alter table test.test_partition_every add partition p20120105_6
start ('2012-01-05'::date) END ('2012-01-07'::date);

　　07,drop/truncate分区

alter table test.test_partition_every drop partition p20120105_6;

alter table test.test_partition_every truncate partition p20120105_6;

　　08,拆分分区

alter table test.test_partition_every split partition p20120105_6
at(('2012-01-06'::date)) into (partition p20120105,partition p20120106);

　　09,交换分区

alter table test.test_partition_every exchange partition p20120102 with table test.test_on_partition;