PostgreSQL PITR实验

磨砺技术珠矶，践行数据之道，追求卓越价值

回到上一级页面： PostgreSQL基础知识与基本操作索引页回到顶级页面：PostgreSQL索引页

看PostgreSQL中与PITR相关的设定：

wal_level=archive

…

# - Archiving -

archive_mode = on                                     # allows archiving to be done

                                                      # (change requires restart)

archive_command = 'cp %p "/usr/local/pgsql/arch/%f"'  # command to use to archive a logfile segment

#archive_timeout = 0                                  # force a logfile segment switch after this

                                                      # number of seconds; 0 disables

开始，在运行之前，看xlog状况：

[postgres@pg201 pgsql]$ ls -lrt ./data/pg_xlog

total

-rw-------  postgres postgres  Aug   :

drwx------  postgres postgres      Aug   : archive_status

-rw-------  postgres postgres  Aug   :

[postgres@pg201 pgsql]$                             

[postgres@pg201 pgsql]$ ls -lrt ./arch

total

-rw-------  postgres postgres  Aug   :

[postgres@pg201 pgsql]$

为文件拷贝进行准备：

[postgres@pg201 pgsql]$ ./bin/psql

psql (9.1.2)

Type "help" for help.                

postgres=# select pg_start_backup('gao');

 pg_start_backup

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

 0/3000020

(1 row)                

postgres=#

此时，xlog和archive log状况如下：

[postgres@pg201 pgsql]$

[postgres@pg201 pgsql]$ ls -lrt ./data/pg_xlog

total 49216

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000004

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

drwx------ 2 postgres postgres     4096 Aug  6 14:17 archive_status

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000003                            

[postgres@pg201 pgsql]$ ls -lrt ./arch

total 32808

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000001

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

[postgres@pg201 pgsql]$

完成文件拷贝：

[postgres@pg201 pgsql]$ tar -cvf basedata.tar ./data

结束基础备份：

postgres=# select pg_stop_backup();

NOTICE:  pg_stop_backup complete, all required WAL segments have been archived

 pg_stop_backup

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

 0/30000A0

(1 row)                            

postgres=#

再次查看log:

[postgres@pg201 pgsql]$ ls -lrt ./data/pg_xlog

total 49220

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000004

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

drwx------ 2 postgres postgres     4096 Aug  6 14:20 archive_status

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000003

[postgres@pg201 pgsql]$                                 

[postgres@pg201 pgsql]$ ls -lrt ./arch

total 49216

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000001

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000003

[postgres@pg201 pgsql]$

可以看到产生了一个backup文件，看内容：

[postgres@pg201 pgsql]$ cat ./data/pg_xlog/000000010000000000000003.00000020.backup

START WAL LOCATION: 0/3000020 (file 000000010000000000000003)

STOP WAL LOCATION: 0/30000A0 (file 000000010000000000000003)

CHECKPOINT LOCATION: 0/3000020

BACKUP METHOD: pg_start_backup

START TIME: 2013-08-06 14:17:53 CST

LABEL: gao

STOP TIME: 2013-08-06 14:20:53 CST

[postgres@pg201 pgsql]$

强制日志切换，模拟大量数据产生导致日志switch:

postgres=# select pg_switch_xlog();

 pg_switch_xlog

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

 0/4000000

(1 row)            

postgres=#

log发生了变化：

[postgres@pg201 pgsql]$ ls -lrt ./data/pg_xlog

total 49220

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000006

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:25 000000010000000000000004

drwx------ 2 postgres postgres     4096 Aug  6 14:27 archive_status

-rw------- 1 postgres postgres 16777216 Aug  6 14:27 000000010000000000000005

[postgres@pg201 pgsql]$

[postgres@pg201 pgsql]$ ls -lrt ./arch

total 65620

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000001

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000003

-rw------- 1 postgres postgres 16777216 Aug  6 14:25 000000010000000000000004

[postgres@pg201 pgsql]$

再作几次log switch，最后是这样：

[postgres@pg201 pgsql]$ ls -lrt ./data/pg_xlog

total 49220

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000006

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:25 000000010000000000000004

drwx------ 2 postgres postgres     4096 Aug  6 14:27 archive_status

-rw------- 1 postgres postgres 16777216 Aug  6 14:27 000000010000000000000005

[postgres@pg201 pgsql]$

[postgres@pg201 pgsql]$ ls -lrt ./arch

total 65620

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000001

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000003

-rw------- 1 postgres postgres 16777216 Aug  6 14:25 000000010000000000000004

[postgres@pg201 pgsql]$

经过一次正常关机后，日志再次变化：

[postgres@pg201 pgsql]$ ./bin/pg_ctl -D ./data stop

日志变为：

[postgres@pg201 pgsql]$ ls -lrt ./data/pg_xlog

total 49220

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:25 000000010000000000000007

drwx------ 2 postgres postgres     4096 Aug  6 14:39 archive_status

-rw------- 1 postgres postgres 16777216 Aug  6 14:39 000000010000000000000006

-rw------- 1 postgres postgres 16777216 Aug  6 14:39 000000010000000000000005

[postgres@pg201 pgsql]$

[postgres@pg201 pgsql]$ ls -lrt ./arch

total 82024

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000001

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000003

-rw------- 1 postgres postgres 16777216 Aug  6 14:25 000000010000000000000004

-rw------- 1 postgres postgres 16777216 Aug  6 14:39 000000010000000000000005

[postgres@pg201 pgsql]$

重新启动PostgreSQL，添加数据，然后杀掉进程，模拟崩溃：

[postgres@pg201 pgsql]$ ./bin/psql

psql (9.1.2)

Type "help" for help.

postgres=# create table test(id integer);

CREATE TABLE

postgres=# insert into test values(1);

INSERT 0 1

postgres=# \q

[postgres@pg201 pgsql]$

[root@pg201 ~]# ps -ef | grep post

root      3120  3049  0 14:08 pts/1    00:00:00 su - postgres

postgres  3121  3120  0 14:08 pts/1    00:00:00 -bash

root      3216  3191  0 14:15 pts/2    00:00:00 su - postgres

postgres  3217  3216  0 14:15 pts/2    00:00:00 -bash

root      3299  3279  0 14:17 pts/3    00:00:00 su - postgres

postgres  3300  3299  0 14:17 pts/3    00:00:00 -bash

postgres  3516     1  0 15:00 pts/2    00:00:00 /usr/local/pgsql/bin/postgres -D ./data

postgres  3518  3516  0 15:00 ?        00:00:00 postgres: writer process

postgres  3519  3516  0 15:00 ?        00:00:00 postgres: wal writer process

postgres  3520  3516  0 15:00 ?        00:00:00 postgres: autovacuum launcher process

postgres  3521  3516  0 15:00 ?        00:00:00 postgres: archiver process

postgres  3522  3516  0 15:00 ?        00:00:00 postgres: stats collector process

root      3559  3535  0 15:01 pts/4    00:00:00 grep post

[root@pg201 ~]# kill -s SIGQUIT 3516

[root@pg201 ~]#

下面开始回复动作、找出基础备份：

[postgres@pg201 pgsql]$ tar -xvf basedata.tar ./data

其pg_xlog目录下，还是旧的online redo log:

[postgres@pg201 pg_xlog]$ ls -lrt

total 49216

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000004

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

drwx------ 2 postgres postgres     4096 Aug  6 14:17 archive_status

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000003

[postgres@pg201 pg_xlog]$

把它清空后，不补足online redolog，仅仅用 archive log的效果：

在解压好的data目录下，建立 recovery.conf文件：

[postgres@pg201 data]$ cat recovery.conf

restore_command = 'cp /usr/local/pgsql/arch/%f %p'

[postgres@pg201 data]$

启动PostgreSQL：

[postgres@pg201 pgsql]$ ./bin/pg_ctl -D ./data status

pg_ctl: no server running

[postgres@pg201 pgsql]$ ./bin/pg_ctl -D ./data start

pg_ctl: another server might be running; trying to start server anyway

server starting

[postgres@pg201 pgsql]$ LOG:  database system was interrupted; last known up at 2013-08-06 14:17:53 CST

LOG:  creating missing WAL directory "pg_xlog/archive_status"

LOG:  starting archive recovery

LOG:  restored log file "000000010000000000000003" from archive

LOG:  redo starts at 0/3000078

LOG:  consistent recovery state reached at 0/4000000

LOG:  restored log file "000000010000000000000004" from archive

LOG:  restored log file "000000010000000000000005" from archive

cp: cannot stat `/usr/local/pgsql/arch/000000010000000000000006': No Such File or Directory

LOG:  could not open file "pg_xlog/000000010000000000000006" (log file 0, segment 6): No Such File or Directory

LOG: redo done at 0/5000078 LOG: restored log file "000000010000000000000005" from archive cp: cannot stat `/usr/local/pgsql/arch/00000002.history': No Such File or Directory

LOG: selected new timeline ID:  cp: cannot stat `/usr/local/pgsql/arch/00000001.history':No Such File or Directory

LOG: archive recovery complete 
LOG: autovacuum launcher started 
LOG: database system is ready to accept connections

因为我的创建table test 的动作虽然计入了 redo log中，但是由于突然的崩溃，没有来得及放入archive 日志中，所以，没有得到恢复：

[postgres@pg201 pgsql]$ ./bin/psql

psql (9.1.2)

Type "help" for help.

postgres=# select * from test;

ERROR:  relation "test" does not exist

LINE 1: select * from test;

                      ^

postgres=#

此时，看Log，就可以清晰地看到TimeLine的痕迹了：

[postgres@pg201 pgsql]$ ls -lrt ./data/pg_xlog

total 49220

-rw------- 1 postgres postgres       56 Aug  6 15:11 00000002.history

-rw------- 1 postgres postgres 16777216 Aug  6 15:11 00000000000000000000005

-rw------- 1 postgres postgres 16777216 Aug  6 15:17 00000000000000000000006

drwx------ 2 postgres postgres     4096 Aug  6 15:17 archive_status

-rw------- 1 postgres postgres 16777216 Aug  6 15:17 00000000000000000000007

[postgres@pg201 pgsql]$

[postgres@pg201 pgsql]$ ls -lrt ./arch

total 114836

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000001

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

-rw------- 1 postgres postgres      267 Aug  6 14:20 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 14:20 000000010000000000000003

-rw------- 1 postgres postgres 16777216 Aug  6 14:25 000000010000000000000004

-rw------- 1 postgres postgres 16777216 Aug  6 14:39 000000010000000000000005

-rw------- 1 postgres postgres       56 Aug  6 15:11 0000000.history

-rw------- 1 postgres postgres 16777216 Aug  6 15:17 00000000000000000000006

-rw------- 1 postgres postgres 16777216 Aug  6 15:17 00000000000000000000005

[postgres@pg201 pgsql]$ cat ./arch/00000002.history

1       000000010000000000000005        no recovery target specified

[postgres@pg201 pgsql]$

这次再看清空pg_xlog目录后，补足online redolog+archive log的效果：

还要把 pg_xlog/archive_status 目录清空：

[postgres@pg201 pgsql]$ cp ./data.bak/pg_xlog/*  ./data/pg_xlog/

...

[postgres@pg201 pg_xlog]$ ls -lrt

total 98432

-rw------- 1 postgres postgres 16777216 Aug  6 14:16 000000010000000000000004

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000002

-rw------- 1 postgres postgres 16777216 Aug  6 14:17 000000010000000000000003

-rw------- 1 postgres postgres      267 Aug  6 15:41 000000010000000000000003.00000020.backup

-rw------- 1 postgres postgres 16777216 Aug  6 15:41 000000010000000000000007

-rw------- 1 postgres postgres 16777216 Aug  6 15:41 000000010000000000000006

-rw------- 1 postgres postgres 16777216 Aug  6 15:41 000000010000000000000005

drwx------ 2 postgres postgres     4096 Aug  6 15:44 archive_status

[postgres@pg201 pg_xlog]$

此时，启动 PostgreSQL:

[postgres@pg201 pgsql]$ ./bin/pg_ctl -D ./data start

pg_ctl: another server might be running; trying to start server anyway

server starting

[postgres@pg201 pgsql]$ LOG:  database system was interrupted; last known up at 2013-08-06 14:17:53 CST

LOG:  starting archive recovery

LOG:  restored log file "000000010000000000000003" from archive

LOG:  redo starts at 0/3000078

LOG:  consistent recovery state reached at 0/4000000

LOG:  restored log file "000000010000000000000004" from archive

LOG:  restored log file "000000010000000000000005" from archive

cp: cannot stat `/usr/local/pgsql/arch/000000010000000000000006': No such file or directory

LOG:  record with zero length at 0/6011770

LOG:  redo done at 0/6011730

LOG:  last completed transaction was at log time 2013-08-06 15:01:17.757945+08

LOG:  restored log file "00000002.history" from archive

cp: cannot stat `/usr/local/pgsql/arch/00000003.history': No such file or directory

LOG:  selected new timeline ID: 3

cp: cannot stat `/usr/local/pgsql/arch/00000001.history': No such file or directory

LOG:  archive recovery complete

LOG:  autovacuum launcher started

LOG:  database system is ready to accept connections

看数据：

[postgres@pg201 pgsql]$ ./bin/psql

psql (9.1.2)

Type "help" for help.

postgres=# select * from test;

 id

----

  1

(1 row)

postgres=#

回到上一级页面： PostgreSQL基础知识与基本操作索引页回到顶级页面：PostgreSQL索引页

磨砺技术珠矶，践行数据之道，追求卓越价值

PostgreSQL PITR实验的更多相关文章

PostgreSQL基础知识与基本操作索引页
磨砺技术珠矶,践行数据之道,追求卓越价值返回顶级页:PostgreSQL索引页 luckyjackgao@gmail.com 本页记录所有本人所写的PostgreSQL的基础知识和基本操作相关文摘和 ...
PostgreSQL的PITR中，对 unfilled wal log 如何处理为好
磨砺技术珠矶,践行数据之道,追求卓越价值回到上一级页面: PostgreSQL基础知识与基本操作索引页回到顶级页面:PostgreSQL索引页通过实验,可以发现,PostgreSQL中使 ...
PostgreSQL LIKE 查询效率提升实验<转>
一.未做索引的查询效率作为对比,先对未索引的查询做测试 EXPLAIN ANALYZE select * from gallery_map where author = '曹志耘'; QUERY P ...
PostgreSQL数据库表名的大小写实验
磨砺技术珠矶,践行数据之道,追求卓越价值回到上一级页面:PostgreSQL基础知识与基本操作索引页回到顶级页面:PostgreSQL索引页[作者高健@博客园 luckyjackgao@g ...
postgresql基于备份点PITR恢复
实验目的: 01.基于备份点直接恢复数据库 02.基于备份点后续增量wal日志恢复到特定的时间点实验环境: centos7 postgresql9.5 01.安装postgresql9.5 post ...
PostgreSQL的 PITR实战---运用 recovery_target_time
磨砺技术珠矶,践行数据之道,追求卓越价值回到上一级页面: PostgreSQL基础知识与基本操作索引页回到顶级页面:PostgreSQL索引页看了很多的例子,没有发现具体讲 recove ...
PostgreSQL的xlog实验一
磨砺技术珠矶,践行数据之道,追求卓越价值回到上一级页面: PostgreSQL基础知识与基本操作索引页回到顶级页面:PostgreSQL索引页一,从initdb开始,initdb执行结束 ...
PostgreSQL设置事务隔离级别实验
apple=# begin; BEGIN apple=# set transaction ISOLATION LEVEL read committed ; SET apple=# select * f ...
对PostgreSQL数据库的hstore类型建立GisT索引的实验
磨砺技术珠矶,践行数据之道,追求卓越价值回到上一级页面:PostgreSQL基础知识与基本操作索引页回到顶级页面:PostgreSQL索引页[作者高健@博客园 luckyjackgao@g ...

随机推荐

leetcode 之 Degree of an Array
1.题目描述 Given a non-empty array of non-negative integers nums, the degree of this array is defined as ...
ARM CPU 架构
1978年12月5日,物理学家赫尔曼·豪泽(Hermann Hauser)和工程师Chris Curry,在英国剑桥创办了CPU公司(Cambridge Processing Unit),主要业务是为 ...
Oracle EBS 获取公司段的本位币
SELECT gls.currency_code FROM hr_organization_information_v t, gl_sets_of_books gls WHERE t.org_info ...
用TableView写带特效的cell
用TableView写带特效的cell 效果: 源码地址: https://github.com/YouXianMing/UI-Component-Collection 分析: 在UIScrollVi ...
定制选择范围的按钮RangeButton
定制选择范围的按钮RangeButton 效果: 源码: RangeButton.h 与 RangeButton.m // // RangeButton.h // PulsingView // // ...
产生渐变色的view
产生渐变色的view 效果源码 https://github.com/YouXianMing/UI-Component-Collection // // GradientColorView.h // ...
基于scrapyd爬虫发布总结
一.版本情况 python以丰富的三方类库取得了众多程序员的认可,但也因此带来了众多的类库版本问题,本文总结的内容是基于最新的类库版本. 1.scrapy版本:1.1.0 D:\python\Spid ...
Windows批量修改文件夹及子文件夹下文件的扩展名
实例:将 D:/backup 目录下所有后缀名为 “.zip” 的文件替换为 “.exe” 后缀 bat批处理: @echo off rem 不显示执行过程 D: rem 切换至指定盘符 cd D:/ ...
成为技术领导者笔记--领导的MOI模型
一. 想让周围环境发生变化,环境必须包含三个条件: M:激励(Motivation)---有奖品或有困难,这样才对相关人员有推动力或吸引力. O:组织(organization)---利用现有的组织结 ...
Anaconda 包管理工具 conda 进行虚拟环境管理入门
在基于 python 进行数据分析.机器学习等领域的实践和学习时,由于代码的更迭和更新,运行他人实现的代码或尝试安装新的工具库时往往需要指定特定版本的其他工具库,以满足特定环境的构建条件.而将同一工具 ...

PostgreSQL PITR实验

PostgreSQL PITR实验的更多相关文章

随机推荐

热门专题