1 建立分区

1.1. 创建主表

CREATE TABLE measurement (

city_id int not null,

logdate date not null,

peaktemp int,

unitsales int

);

CREATE TABLE

这里date类型精确到天，如'2006-02-01'，如果时间精确到秒，如'2013-01-09 00:00:00' 使用timestamp 类型

1.2. 创建分区表

CREATE TABLEmeasurement_y2006m01 (

CHECK (logdate >= DATE '2006-01-01' AND logdate < DATE '2006-02-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m02 (

CHECK (logdate >= DATE '2006-02-01' AND logdate < DATE '2006-03-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m03 (

CHECK (logdate >= DATE '2006-03-01' AND logdate < DATE '2006-04-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m04 (

CHECK (logdate >= DATE '2006-04-01' AND logdate < DATE '2006-05-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m05 (

CHECK (logdate >= DATE '2006-05-01' AND logdate < DATE '2006-06-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m06 (

CHECK (logdate >= DATE '2006-06-01' AND logdate < DATE '2006-07-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m07 (

CHECK (logdate >= DATE '2006-07-01' AND logdate < DATE '2006-08-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m08 (

CHECK (logdate >= DATE '2006-08-01' AND logdate < DATE '2006-09-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m09 (

CHECK (logdate >= DATE '2006-09-01' AND logdate < DATE '2006-10-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m10 (

CHECK (logdate >= DATE '2006-10-01' AND logdate < DATE '2006-11-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m11 (

CHECK (logdate >= DATE '2006-11-01' AND logdate < DATE '2006-12-01' )

) INHERITS (measurement);

CREATE TABLE measurement_y2006m12 (

CHECK (logdate >= DATE '2006-12-01' AND logdate < DATE '2007-01-01' )

) INHERITS (measurement);

CREATE TABLE

1.3. 分区键上建索引

CREATE INDEX measurement_y2006m02_logdateON measurement_y2006m02 (logdate);

CREATE INDEX measurement_y2006m03_logdateON measurement_y2006m03 (logdate);

CREATE INDEX measurement_y2006m04_logdateON measurement_y2006m04 (logdate);

CREATE INDEX measurement_y2006m05_logdateON measurement_y2006m05 (logdate);

CREATE INDEX measurement_y2006m06_logdateON measurement_y2006m06 (logdate);

CREATE INDEX measurement_y2006m07_logdateON measurement_y2006m07 (logdate);

CREATE INDEX measurement_y2006m08_logdateON measurement_y2006m08 (logdate);

CREATE INDEX measurement_y2006m09_logdateON measurement_y2006m09 (logdate);

CREATE INDEX measurement_y2006m10_logdateON measurement_y2006m10 (logdate);

CREATE INDEX measurement_y2006m11_logdateON measurement_y2006m11 (logdate);

CREATE INDEX measurement_y2006m12_logdateON measurement_y2006m12 (logdate);

CREATE INDEX

执行insert into measurement命令向表中插入数据，希望数据按一定的规则插入相应分区，需要前期人工处理下，这里有两种策略，建立触发器和RULE规则。建立触发器方法需要1.4、1.5两步操作，1.6是建立RULE规则方法。

1.4. 创建触发器函数

CREATE OR REPLACEFUNCTION measurement_insert_trigger()

RETURNSTRIGGER AS $$

BEGIN

IF ( NEW.logdate >= DATE '2006-01-01'AND

NEW.logdate < DATE '2006-02-01' )THEN

INSERT INTO measurement_y2006m01 VALUES(NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-02-01' AND

NEW.logdate < DATE '2006-03-01' ) THEN

INSERT INTO measurement_y2006m02 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-03-01'AND

NEW.logdate < DATE '2006-04-01' ) THEN

INSERT INTO measurement_y2006m03 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-04-01'AND

NEW.logdate < DATE '2006-05-01' ) THEN

INSERT INTO measurement_y2006m04 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-05-01'AND

NEW.logdate < DATE '2006-06-01' ) THEN

INSERT INTO measurement_y2006m05 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-06-01'AND

NEW.logdate < DATE '2006-07-01' ) THEN

INSERT INTO measurement_y2006m06 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-07-01'AND

NEW.logdate < DATE '2006-08-01' ) THEN

INSERT INTO measurement_y2006m07 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-08-01'AND

NEW.logdate < DATE '2006-09-01' ) THEN

INSERT INTO measurement_y2006m08 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-09-01' AND

NEW.logdate < DATE '2006-10-01' ) THEN

INSERT INTO measurement_y2006m09 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-10-01' AND

NEW.logdate < DATE '2006-11-01' ) THEN

INSERT INTO measurement_y2006m10 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-11-01'AND

NEW.logdate < DATE '2006-12-01' ) THEN

INSERT INTO measurement_y2006m11 VALUES (NEW.*);

ELSIF ( NEW.logdate >= DATE '2006-12-01' AND

NEW.logdate < DATE '2007-01-01' ) THEN

INSERT INTO measurement_y2006m12 VALUES (NEW.*);

ELSE

RAISE EXCEPTION 'Date out of range. Fix the measurement_insert_trigger() function!';

END IF;

RETURN NULL;

END;

LANGUAGE plpgsql;

CREATE FUNCTION

说明：如果不想丢失数据，上面的ELSE 条件可以改成 INSERTINTO measurement _error_ logdate VALUES (NEW.*); 同时需要创建一张结构和measurement 一样的表measurement_error_logdate，这样，错误的logdate 数据就可以插入到这张表中而不是报错了。

1.5. 创建触发器

CREATE TRIGGERinsert_measurement_trigger

BEFORE INSERT ON measurement

FOR EACH ROW EXECUTE PROCEDUREmeasurement_insert_trigger();

CREATE TRIGGER

1.6. 改用RULE规则

也可以不用触发器，改用RULE规则，为每个分表创建规则

CREATE RULE measurement_y2006m01 AS
ON INSERT TO measurement where(NEW.logdate >= DATE '2006-01-01' AND

NEW.logdate < DATE'2006-02-01')

DO INSTEAD

INSERT INTO measurement VALUES (NEW.*);

CREATE RULE measurement_y2006m02 AS
ON INSERT TO measurement where(NEW.logdate >= DATE '2006-02-01' AND

NEW.logdate < DATE'2006-03-01')

DO INSTEAD

INSERT INTO measurement VALUES (NEW.*);

……省略

2 测试

2.1. 查看所有表

mydb=# \d

List of relations

Schema | Name | Type | Owner

--------+----------------------+-------+-------

public | measurement | table | super

public | measurement_y2006m01 | table | super

public | measurement_y2006m02 | table | super

public | measurement_y2006m03 | table | super

public | measurement_y2006m04 | table | super

public | measurement_y2006m05 | table | super

public | measurement_y2006m06 | table | super

public | measurement_y2006m07 | table | super

public | measurement_y2006m08 | table | super

public | measurement_y2006m09 | table | super

public | measurement_y2006m10 | table | super

public | measurement_y2006m11 | table | super

public | measurement_y2006m12 | table | super

(13rows)

2.2. 查看主表结构

mydb=#\d measurement

Table "public.measurement"

Column | Type | Modifiers

-----------+---------+-----------

city_id | integer | not null

logdate | date |not null

peaktemp | integer |

unitsales | integer |

Triggers:

insert_measurement_trigger BEFORE INSERT ON measurement FOR EACH ROWEXECUTE PROCEDURE measurement_insert_trigger().

Number of child tables: 12 (Use \d+ to listthem.)

2.3. 插入数据

mydb=# insert into measurement values(1,date '2006-02-10',1,1);

INSERT 0 0

mydb=# insert into measurement values(1,date '2006-03-10',1,1);

INSERT 0 0

mydb=# insert into measurement values(1,date '2006-04-10',1,1);

INSERT 0 0

2.4. 查看主表数据

mydb=# select * from measurement;

city_id | logdate | peaktemp | unitsales

---------+------------+----------+-----------

1 | 2006-02-10 | 1 | 1

1 | 2006-03-10 | 1 | 1

1 | 2006-04-10 | 1 | 1

(3 rows)

2.5. 查看分表数据

mydb=# select * frommeasurement_y2006m03;

city_id | logdate | peaktemp | unitsales

---------+------------+----------+-----------

1 | 2006-03-10 | 1 | 1

(1 row)

2.6. 分区排除和查询优化

默认情况下，是不会开启分区排除的，索引针对一个基于分区键条件的检索，要扫描所有的分区

mydb=# SET constraint_exclusion = off;

SET

mydb=# EXPLAIN SELECT count(*) FROMmeasurement WHERE logdate <= DATE '2006-02-01';

QUERYPLAN

-----------------------------------------------------------------------------------------------

Aggregate (cost=436.80..436.81 rows=1 width=0)

-> Append (cost=0.00..417.62 rows=7670 width=0)

-> Seq Scan onmeasurement (cost=0.00..32.12 rows=590width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m01 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m02 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m03 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m04 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m05 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan on measurement_y2006m06measurement (cost=0.00..32.12 rows=590width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m07 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <= '2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m08 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m09 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m10 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m11 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan onmeasurement_y2006m12 measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

(28 rows)

2.7. 开启约束排除

mydb=# SET constraint_exclusion = on;

SET

mydb=> EXPLAIN SELECT count(*) FROM measurement WHERE logdate<= DATE '2006-02-01';

QUERYPLAN

-----------------------------------------------------------------------------------------------

Aggregate (cost=100.80..100.81 rows=1 width=0)

-> Append (cost=0.00..96.38 rows=1770 width=0)

-> Seq Scan on measurement (cost=0.00..32.12 rows=590 width=0)

Filter: (logdate <='2006-02-01'::date)

-> Seq Scan on measurement_y2006m01measurement (cost=0.00..32.12 rows=590width=0)

Filter:(logdate <= '2006-02-01'::date)

-> Seq Scan on measurement_y2006m02 measurement (cost=0.00..32.12 rows=590 width=0)

Filter:(logdate <= '2006-02-01'::date)

(8 rows)

至此分区表建立完成

postgresql 分区表创建及测试的更多相关文章

PostgreSQL分区表实现——pg_pathman分区表管理
该博文用于自己学习记录,内容节选自: https://github.com/digoal/blog/blob/master/201610/20161024_01.md pg_pathman 创建分区表 ...
MongoDB与PostgresQL无责任初步测试
PostgresQL一秒能插入多少条记录,MongoDB呢?读取的情况又如何?我写了一些简单的程序,得出了一些简单的数据,贴在这里分享,继续往下阅读前请注意下本文标题中的“无责任”,这表示此测试结果不 ...
通过arcgis在PostgreSQL中创建企业级地理数据库
部署环境: Win7 64位旗舰版软件版本: PostgreSQL-9.1.3-2-windows-x64 Postgis-pg91x64-setup-2.0.6-1 Arcgis 10.1 SP1 ...
批量创建AD测试账号
在现场中,有时候客户会要求做一下AD压力测试,需要批量创建很多AD用户.奉献此代码供各位参考. 1: <# 2: 3: .DESCRIPTION 4: 批量创建AD测试账号 5: 6 ...
soapui中文操作手册(二)----通过您的WSDL请求创建一个测试
1.通过您的WSDL请求创建一个测试点击加号旁边的导航拓展项目树的Web服务,并选择请求: 在SoapUI Pro中,所述请求编辑出现在右边.SoapUI Pro有一个编辑器,它简化了XML的层 ...
eos教程如何创建eos测试账号并且使用scatter插件
EOS代币租赁平台 --- Chintai平台已经在Jungle测试网络上部署了,欢迎大家来体验. 地址见: Chintai 公测版官网是: Chintai 目前测试网络上面需要用到Scatter插 ...
用ArcMap在PostgreSQL中创建要素类需要执行”create enterprise geodatabase”吗
问:用Acmap在PostgreSQL中创建要素类需要执行"create enterprise geodatabase"吗? 关于这个问题,是在为新员工做postgresql培训后 ...
3）创建，测试，发布第一个NET CORE程序
工具:Visual Studio Code 或者 Visual Studio 环境:.NET CORE 2.0 VS Code很强大当然支持netcore的开发,但是我还是选择更熟悉更强大的VS. ...
PostgreSQL SERIAL创建自增列
PostgreSQL SERIAL创建自增列本文我们介绍PostgreSQL SERIAL,并展示如何使用serial类型创建表自增列. PostgreSQL SERIAL伪类型 PostgreSQ ...

随机推荐

AudioManager音频管理器
AudioManager音频管理器提供了如下几种常用方法来控制手机音频: 1.adjustStreamVolume(int StreamType,int direction,int flgs):调整手 ...
C#画线源码
画线 private void Form1_Load(object sender, EventArgs e) { this.Paint += new PaintEventHandler(Form1_P ...
js打印对象（object）
function printObject(obj){//obj = {"cid":"C0","ctext":"区县"}; ...
wp8.1 Study8：页面过渡和主题动画（Page transition and Theme animations）
一.在WP8.1中是有动画(Animation)的: 页面导航(默认为旋转式Turnstile).PointerDown/up(默认是倾斜).页面旋转.MenuFlyout出现等等二.页面过渡(Pa ...
c#读写文本文档-1-用file类
using System; using System.Collections.Generic; using System.Linq; using System.Text; using System.I ...
winform错误提示：窗口类名无效(Window class name is not valid)
winfrom 程序在 xp 操作系统上报错提示窗口类名无效(Window class name is not valid) 解决方法注释 Program类里这句 Application.En ...
Nginx SSL配置过程
1. 在godaddy购买了UCC SSL(最多5个域名)的SSL证书 2. 设置证书 -- 管理 -- 3. 需要制作证书申请CSR文件(在线工具制作或者openssl命令制作),保存CSR和key ...
ODI中显示us7ascii字符集的测试
安装oracle DB时,选择的字符集:美国.英语.US7ASCII. 在不设置nls_lang的情况,插入中文,成功,但存进去的是乱码,select看到也是??(无论后面再怎么设置nls_lang) ...
iOS 7 教程：定制iOS 7中的导航栏和状态栏
目录(?)[-] iOS 7中默认的导航栏设置导航栏的背景颜色在导航栏中使用背景图片定制返回按钮的颜修改导航栏标题的字体修改导航栏标题为图片添加多个按钮修改状态栏的风格隐藏状态栏总结 ...
poj2777 线段树
//Accepted 4768 KB 391 ms //线段树,延时标记的应用 //对于每一段,用一个int表示被着色的情况,change标记该段的颜色是否发生整体的改变,即这一段 //用没用被全部涂 ...

postgresql 分区表创建及测试

1 建立分区

1.1. 创建主表

1.2. 创建分区表

1.3. 分区键上建索引

1.4. 创建触发器函数

1.5. 创建触发器

1.6. 改用RULE规则

2 测试

2.1. 查看所有表

2.2. 查看主表结构

2.3. 插入数据

2.4. 查看主表数据

2.5. 查看分表数据

2.6. 分区排除和查询优化

2.7. 开启约束排除

postgresql 分区表创建及测试的更多相关文章

随机推荐

热门专题