分类目录归档:PostgreSQL恢复

PostgreSQL恢复系列:pg_filedump基本使用

当PostgreSQL遇到重大故障,使用各种方法都无法直接启动数据库,可以考虑使用类似oracle dul工具,直接离线方式读取文件进行恢复.这个工具为pg_filedump
pg_filedump安装

[root@xifenfei ~]# yum install pg_filedump_14.x86_64
Loaded plugins: langpacks, ulninfo
Resolving Dependencies
--> Running transaction check
---> Package pg_filedump_14.x86_64 0:14.1-1.rhel7 will be installed
--> Finished Dependency Resolution

Dependencies Resolved

======================================================================================================================
 Package                            Arch      Version                 Repository                                 Size
======================================================================================================================
Installing:                                 
 pg_filedump_14                     x86_64    14.1-1.rhel7            pgdg14                                     43 k
                                            
Transaction Summary                         
======================================================================================================================
Install  1 Package

Total download size: 43 k
Installed size: 81 k
Is this ok [y/d/N]: y
Downloading packages:
pg_filedump_14-14.1-1.rhel7.x86_64.rpm                                              |  43 kB  00:00:02     
Running transaction check
Running transaction test
Transaction test succeeded
Running transaction
  Installing : pg_filedump_14-14.1-1.rhel7.x86_64                                                      1/1 
  Verifying  : pg_filedump_14-14.1-1.rhel7.x86_64                                                      1/1 

Installed:
  pg_filedump_14.x86_64 0:14.1-1.rhel7                                                                     

Complete!
-bash-4.2$ pg_filedump 

Version 14.1 (for PostgreSQL 8.x .. 14.x)
Copyright (c) 2002-2010 Red Hat, Inc.
Copyright (c) 2011-2022, PostgreSQL Global Development Group

Usage: pg_filedump [-abcdfhikxy] [-R startblock [endblock]] [-D attrlist] [-S blocksize] [-s segsize] [-n segnumber] file

Display formatted contents of a PostgreSQL heap/index/control file
Defaults are: relative addressing, range of the entire file, block
               size as listed on block 0 in the file

The following options are valid for heap and index files:
  -a  Display absolute addresses when formatting (Block header
      information is always block relative)
  -b  Display binary block images within a range (Option will turn
      off all formatting options)
  -d  Display formatted block content dump (Option will turn off
      all other formatting options)
  -D  Decode tuples using given comma separated list of types
      Supported types:
        bigint bigserial bool char charN date float float4 float8 int
        json macaddr name numeric oid real serial smallint smallserial text
        time timestamp timestamptz timetz uuid varchar varcharN xid xml
      ~ ignores all attributes left in a tuple
  -f  Display formatted block content dump along with interpretation
  -h  Display this information
  -i  Display interpreted item details
  -k  Verify block checksums
  -o  Do not dump old values.
  -R  Display specific block ranges within the file (Blocks are
      indexed from 0)
        [startblock]: block to start at
        [endblock]: block to end at
      A startblock without an endblock will format the single block
  -s  Force segment size to [segsize]
  -t  Dump TOAST files
  -v  Ouput additional information about TOAST relations
  -n  Force segment number to [segnumber]
  -S  Force block size to [blocksize]
  -x  Force interpreted formatting of block items as index items
  -y  Force interpreted formatting of block items as heap items

The following options are valid for control files:
  -c  Interpret the file listed as a control file
  -f  Display formatted content dump along with interpretation
  -S  Force block size to [blocksize]
Additional functions:
  -m  Interpret file as pg_filenode.map file and print contents (all
      other options will be ignored)

Report bugs to <pgsql-bugs@postgresql.org>

创建测试表

-bash-4.2$ psql
psql (14.3)
Type "help" for help.

postgres=# create table pg_xifenfei(id int,name varchar(100));
CREATE TABLE
postgres=# insert into pg_xifenfei values(1,'www.xifenfei.com');
INSERT 0 1
postgres=# insert into pg_xifenfei values(2,'xienfei_pg_recovery');
INSERT 0 1
postgres=# select * from pg_xifenfei;
 id |        name         
----+---------------------
  1 | www.xifenfei.com
  2 | xienfei_pg_recovery
(2 rows)

postgres=# 

pg_filedump恢复数据

-bash-4.2$ pg_filedump /var/lib/pgsql/14/data/base/14487/16384

*******************************************************************
* PostgreSQL File/Block Formatted Dump Utility
*
* File: /var/lib/pgsql/14/data/base/14487/16384
* Options used: None
*******************************************************************

Block    0 ********************************************************
<Header> -----
 Block Offset: 0x00000000         Offsets: Lower      32 (0x0020)
 Block: Size 8192  Version    4            Upper    8096 (0x1fa0)
 LSN:  logid      0 recoff 0x16299cf0      Special  8192 (0x2000)
 Items:    2                      Free Space: 8064
 Checksum: 0x0000  Prune XID: 0x00000000  Flags: 0x0000 ()
 Length (including item array): 32

<Data> -----
 Item   1 -- Length:   45  Offset: 8144 (0x1fd0)  Flags: NORMAL
 Item   2 -- Length:   48  Offset: 8096 (0x1fa0)  Flags: NORMAL


*** End of File Encountered. Last Block Read: 0 ***

-bash-4.2$ pg_filedump -D int,charn /var/lib/pgsql/14/data/base/14487/16384|grep COPY
COPY: 1 www.xifenfei.com
COPY: 2 xienfei_pg_recovery
-bash-4.2$ pg_filedump -D int,charn /var/lib/pgsql/14/data/base/14487/16384|grep COPY
> |awk '{$1=null;print $0}'>/tmp/pg_xifenfei_rec
-bash-4.2$ sed -i 's/^[ ]*//g' /tmp/pg_xifenfei_rec

导入数据验证

postgres=# truncate table pg_xifenfei;
TRUNCATE TABLE
postgres=# select * from pg_xifenfei;
 id | name 
----+------
(0 rows)
postgres=# copy pg_xifenfei from '/tmp/pg_xifenfei_rec'(DELIMITER ' ');
COPY 2
postgres=# select * from pg_xifenfei;
 id |        name         
----+---------------------
  1 | www.xifenfei.com
  2 | xienfei_pg_recovery
(2 rows)

通过上述简单测试证明,在PG数据库出现极端情况下,可以使用该方法进行最后的数据恢复,减少因为数据丢失带来的损失.

发表在 PostgreSQL恢复 | 标签为 , , | 评论关闭

PostgreSQL恢复系列:pg_control异常恢复

在PG中pg_control文件类似oracle数据库的control文件(控制文件),在Oracle中如果该文件丢失/损坏,可以通过alter database create controlfile命令进行创建,对于PG数据库来说也可以通过pg_resetwal命令来实现创建,由于pg_control文件损坏,需要人工指定一些参数完成pg_resetwal相关操作
pg_resetwal 使用说明

-bash-4.2$ pg_resetwal --help
pg_resetwal resets the PostgreSQL write-ahead log.

Usage:
  pg_resetwal [OPTION]... DATADIR

Options:
  -c, --commit-timestamp-ids=XID,XID
                                   set oldest and newest transactions bearing
                                   commit timestamp (zero means no change)
 [-D, --pgdata=]DATADIR            data directory
  -e, --epoch=XIDEPOCH             set next transaction ID epoch
  -f, --force                      force update to be done
  -l, --next-wal-file=WALFILE      set minimum starting location for new WAL
  -m, --multixact-ids=MXID,MXID    set next and oldest multitransaction ID
  -n, --dry-run                    no update, just show what would be done
  -o, --next-oid=OID               set next OID
  -O, --multixact-offset=OFFSET    set next multitransaction offset
  -u, --oldest-transaction-id=XID  set oldest transaction ID
  -V, --version                    output version information, then exit
  -x, --next-transaction-id=XID    set next transaction ID
      --wal-segsize=SIZE           size of WAL segments, in megabytes
  -?, --help                       show this help, then exit

Report bugs to <pgsql-bugs@lists.postgresql.org>.
PostgreSQL home page: <https://www.postgresql.org/>

确认现在业务表记录情况

-bash-4.2$ psql
psql (14.3)
Type "help" for help.

postgres=# select count(1) from ac_event;
 count  
--------
 246266
(1 row)

模拟pg_control文件异常

-bash-4.2$ ps -ef|grep postgres
postgres  37178      1  0 09:58 ?        00:00:00 /usr/pgsql-14/bin/postgres -D /var/lib/pgsql/14/data
postgres  37179  37178  0 09:58 ?        00:00:00 postgres: logger 
postgres  37181  37178  0 09:58 ?        00:00:00 postgres: checkpointer 
postgres  37182  37178  0 09:58 ?        00:00:00 postgres: background writer 
postgres  37183  37178  0 09:58 ?        00:00:00 postgres: walwriter 
postgres  37184  37178  0 09:58 ?        00:00:00 postgres: autovacuum launcher 
postgres  37185  37178  0 09:58 ?        00:00:00 postgres: stats collector 
postgres  37186  37178  0 09:58 ?        00:00:00 postgres: logical replication launcher 
root      41368  41314  0 11:06 pts/1    00:00:00 su - postgres
postgres  41369  41368  0 11:06 pts/1    00:00:00 -bash
postgres  45071  41369  0 12:07 pts/1    00:00:00 ps -ef
postgres  45072  41369  0 12:07 pts/1    00:00:00 grep --color=auto postgres
-bash-4.2$ kill -9 37178
-bash-4.2$ ps -ef|grep postgres
root      41368  41314  0 11:06 pts/1    00:00:00 su - postgres
postgres  41369  41368  0 11:06 pts/1    00:00:00 -bash
postgres  45095  41369  0 12:08 pts/1    00:00:00 ps -ef
postgres  45096  41369  0 12:08 pts/1    00:00:00 grep --color=auto postgres
-bash-4.2$ pwd
/var/lib/pgsql/14/data/global
-bash-4.2$ ls -l pg_control 
-rw-------. 1 postgres postgres 8192 May 30 12:04 pg_control
-bash-4.2$ rm -rf pg_control 
-bash-4.2$ ls -l pg_control 
ls: cannot access pg_control: No such file or directory

PG启动失败

-bash-4.2$ pg_ctl start 
pg_ctl: another server might be running; trying to start server anyway
waiting for server to start....postgres: could not find the database system
Expected to find it in the directory "/var/lib/pgsql/14/data",
but could not open file "/var/lib/pgsql/14/data/global/pg_control": No such file or directory
 stopped waiting
pg_ctl: could not start server
Examine the log output.

创建空pg_control文件启动依旧失败

-bash-4.2$ touch /var/lib/pgsql/14/data/global/pg_control
-bash-4.2$ pg_ctl start 
pg_ctl: another server might be running; trying to start server anyway
waiting for server to start....2022-05-30 12:09:43.953 CST [45215] PANIC:  
   could not read file "global/pg_control": read 0 of 296
 stopped waiting
pg_ctl: could not start server
Examine the log output.

设置next-wal-file
-l, –next-wal-file=WALFILE,这个参数设置下一个新的WAL文件的最小值,这个值可以从$PGDATA/pg_wal目录下去看最后一个WAL 文件,这个文件的id+1即可

-bash-4.2$ pwd
/var/lib/pgsql/14/data/pg_wal
-bash-4.2$ ls -l
total 16384
-rw-------. 1 postgres postgres 16777216 May 30 12:04 000000010000000000000014
drwx------. 2 postgres postgres        6 May 24 02:20 archive_status
-bash-4.2$ 

这个文件+1,-l 000000010000000000000015
设置next-transaction
-x, –next-transaction-id=XID,这个参数设置pg_control中的下一个XID的值,这个值可以从pg_xact目录下的文件中查询

-bash-4.2$ pwd
/var/lib/pgsql/14/data/pg_xact
-bash-4.2$ ls -ltr
total 8
-rw-------. 1 postgres postgres 8192 May 30 12:03 0000

最后一个是0000,那么下一个XID就是0001,然后乘以 1048576 (0×100000),实际上后面直接加5个0就行了。注意,这个值是16进制的。-x 0×000100000
multixact-ids设置
-m, –multixact-ids=MXID1,MXID2,这个参数包含两个部分,MXID1和MXID2,都可以从$PGDATA/pg_multixact/offsets目录下获得。MXID1的值,首先找到最大值,+1,再乘以 65536 (0×10000,相当于后面加4个0)作为这个参数的前半部分。找到最小的值,后面加4个0,作为MXID2的值

-bash-4.2$ pwd
/var/lib/pgsql/14/data/pg_multixact/offsets
-bash-4.2$ ls -ltr
total 8
-rw-------. 1 postgres postgres 8192 May 29 22:06 0000
-bash-4.2$ 

-m 0×00010000, 0×00000000(由于oldest multitransaction ID不能为0,因此后续这个值需要适当调整)
multixact-offset设置
-O, –multixact-offset=OFFSET,这个参数可以从$PGDATA/pg_multixact/members目录下获得。找到最大值,+1,乘以 52352 (0xCC80)

-bash-4.2$ pwd
/var/lib/pgsql/14/data/pg_multixact/members
-bash-4.2$ ls -ltr
total 8
-rw-------. 1 postgres postgres 8192 May 24 02:20 0000

-O 0xCC80
尝试执行pg_resetwal

-bash-4.2$ pg_resetwal -l 000000010000000000000015 -x 0x000100000 -m 0x00010000,0x00000000 -O 0xCC80 $PGDATA
pg_resetwal: error: oldest multitransaction ID (-m) must not be 0

multixact-ids值不对,进行调整后处理
postmaster.pid文件需要清理
由于PG库异常关闭,需要人工清理掉该文件

-bash-4.2$ pg_resetwal -l 000000010000000000000015 -x 0x000100000 -m 0x00020000,0x00010000 -O 0xCC80 $PGDATA
pg_resetwal: error: lock file "postmaster.pid" exists
-bash-4.2$ rm -rf postmaster.pid 

pg_resetwal结果预览

-bash-4.2$ pg_resetwal -l 000000010000000000000015 -x 0x000100000 -m 0x00020000,0x00010000 -O 0xCC80 $PGDATA
pg_resetwal: warning: pg_control exists but is broken or wrong version; ignoring it
Guessed pg_control values:

pg_control version number:            1300
Catalog version number:               202107181
Database system identifier:           7103392535324046312
Latest checkpoint's TimeLineID:       1
Latest checkpoint's full_page_writes: off
Latest checkpoint's NextXID:          0:3
Latest checkpoint's NextOID:          12000
Latest checkpoint's NextMultiXactId:  1
Latest checkpoint's NextMultiOffset:  0
Latest checkpoint's oldestXID:        3
Latest checkpoint's oldestXID's DB:   0
Latest checkpoint's oldestActiveXID:  0
Latest checkpoint's oldestMultiXid:   1
Latest checkpoint's oldestMulti's DB: 0
Latest checkpoint's oldestCommitTsXid:0
Latest checkpoint's newestCommitTsXid:0
Maximum data alignment:               8
Database block size:                  8192
Blocks per segment of large relation: 131072
WAL block size:                       8192
Bytes per WAL segment:                16777216
Maximum length of identifiers:        64
Maximum columns in an index:          32
Maximum size of a TOAST chunk:        1996
Size of a large-object chunk:         2048
Date/time type storage:               64-bit integers
Float8 argument passing:              by value
Data page checksum version:           0


Values to be changed:

First log segment after reset:        000000010000000000000015
NextMultiXactId:                      131072
OldestMultiXid:                       65536
OldestMulti's DB:                     0
NextMultiOffset:                      52352
NextXID:                              1048576
OldestXID:                            3
OldestXID's DB:                       0

If these values seem acceptable, use -f to force reset.

pg_resetwal进行创建pg_control并启动PG

-bash-4.2$ pg_resetwal -l 000000010000000000000015 -x 0x000100000 -m 0x00020000,0x00010000 -O 0xCC80 -f $PGDATA
pg_resetwal: warning: pg_control exists but is broken or wrong version; ignoring it
Write-ahead log reset
-bash-4.2$ pg_ctl start
waiting for server to start....2022-05-30 13:33:28.266 CST [51437] LOG:  
redirecting log output to logging collector process
2022-05-30 13:33:28.266 CST [51437] HINT:  Future log output will appear in directory "log".
 done
server started

验证数据

-bash-4.2$ psql
psql (14.3)
Type "help" for help.

postgres=#  select count(1) from ac_event;
 count  
--------
 245275
(1 row)

这种方法恢复之后,建议理解dump数据,然后导入到新库中

发表在 PostgreSQL恢复 | 标签为 , , , | 评论关闭

PostgreSQL恢复系列:wal日志丢失恢复

WAL是Write Ahead Log的简写,和oracle的redo日志类似,存放在$PGDATA/pg_xlog中,10版本以后在$PGDATA/pg_wal目录.在oracle数据库中,如果redo丢失,分为active/current和inactive的redo,分别有不同的处理方式,对于oracle需要实例恢复的redo丢失,需要屏蔽数据库一致性,强制打开数据库,对于PG数据库这部分日志丢失该如何恢复,主要是通过pg_resetwal/pg_resetxlog(10以前版本)命令来实现,这里通过一个测试来验证
创建测试表并强制kill数据库

-bash-4.2$ psql
psql (14.3)
Type "help" for help.

postgres=# create table t_xifenfei as select * from pg_database;
SELECT 4
postgres=# select count(1) from t_xifenfei;
 count 
-------
     4
(1 row)

postgres=# \q
-bash-4.2$ ps -ef|grep post
root       1819      1  0 May28 ?        00:00:00 /usr/libexec/postfix/master -w
postfix    1838   1819  0 May28 ?        00:00:00 qmgr -l -t unix -u
postgres  11102      1  0 05:49 ?        00:00:00 /usr/pgsql-14/bin/postgres -D /var/lib/pgsql/14/data
postgres  11103  11102  0 05:49 ?        00:00:00 postgres: logger 
postgres  11105  11102  0 05:49 ?        00:00:00 postgres: checkpointer 
postgres  11106  11102  0 05:49 ?        00:00:00 postgres: background writer 
postgres  11107  11102  0 05:49 ?        00:00:00 postgres: walwriter 
postgres  11108  11102  0 05:49 ?        00:00:00 postgres: autovacuum launcher 
postgres  11109  11102  0 05:49 ?        00:00:01 postgres: stats collector 
postgres  11110  11102  0 05:49 ?        00:00:00 postgres: logical replication launcher 
root      22743  22300  0 18:26 pts/3    00:00:00 su - postgres
postgres  22744  22743  0 18:26 pts/3    00:00:00 -bash
postgres  22937  22744  0 18:28 pts/3    00:00:00 psql
postgres  22938  11102  0 18:28 ?        00:00:00 postgres: postgres postgres [local] idle
postfix   32623   1819  0 21:10 ?        00:00:00 pickup -l -t unix -u
root      33032  32912  0 21:15 pts/2    00:00:00 su - postgres
postgres  33033  33032  0 21:15 pts/2    00:00:00 -bash
postgres  35210  33033  0 21:51 pts/2    00:00:00 ps -ef
postgres  35211  33033  0 21:51 pts/2    00:00:00 grep --color=auto post
-bash-4.2$ kill -9 11102

删除wal日志

-bash-4.2$ pwd
/var/lib/pgsql/14/data/pg_wal
-bash-4.2$ ls -ltr
total 311296
drwx------. 2 postgres postgres        6 May 24 02:20 archive_status
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000014
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000015
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000016
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000017
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000018
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000019
-rw-------. 1 postgres postgres 16777216 May 28 21:29 00000001000000000000001A
-rw-------. 1 postgres postgres 16777216 May 28 21:29 00000001000000000000001B
-rw-------. 1 postgres postgres 16777216 May 28 21:29 00000001000000000000001C
-rw-------. 1 postgres postgres 16777216 May 28 21:29 00000001000000000000001D
-rw-------. 1 postgres postgres 16777216 May 28 21:29 00000001000000000000001E
-rw-------. 1 postgres postgres 16777216 May 28 21:29 00000001000000000000001F
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000020
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000021
-rw-------. 1 postgres postgres 16777216 May 28 21:29 000000010000000000000022
-rw-------. 1 postgres postgres 16777216 May 28 21:30 000000010000000000000023
-rw-------. 1 postgres postgres 16777216 May 28 21:30 000000010000000000000024
-rw-------. 1 postgres postgres 16777216 May 28 21:30 000000010000000000000025
-rw-------. 1 postgres postgres 16777216 May 29 21:51 000000010000000000000013
-bash-4.2$ rm -rf 0000000100000000000000*
-bash-4.2$ ls
archive_status

查询当时数据库需要的最小wal记录

-bash-4.2$ pg_controldata 
pg_control version number:            1300
Catalog version number:               202107181
Database system identifier:           7100998319216817119
Database cluster state:               in production
pg_control last modified:             Sat 28 May 2022 09:36:11 PM CST
Latest checkpoint location:           0/13692F80
Latest checkpoint's REDO location:    0/13692F48
Latest checkpoint's REDO WAL file:    000000010000000000000013   <===需要的记录
Latest checkpoint's TimeLineID:       1
Latest checkpoint's PrevTimeLineID:   1
Latest checkpoint's full_page_writes: on
Latest checkpoint's NextXID:          0:17824
Latest checkpoint's NextOID:          32769
Latest checkpoint's NextMultiXactId:  1
Latest checkpoint's NextMultiOffset:  0
Latest checkpoint's oldestXID:        727
Latest checkpoint's oldestXID's DB:   1
Latest checkpoint's oldestActiveXID:  17824
Latest checkpoint's oldestMultiXid:   1
Latest checkpoint's oldestMulti's DB: 1
Latest checkpoint's oldestCommitTsXid:0
Latest checkpoint's newestCommitTsXid:0
Time of latest checkpoint:            Sat 28 May 2022 09:31:41 PM CST

尝试启动PG

-bash-4.2$ pg_ctl start
pg_ctl: another server might be running; trying to start server anyway
waiting for server to start....2022-05-29 21:52:22.926 CST [35270] LOG:  
redirecting log output to logging collector process
2022-05-29 21:52:22.926 CST [35270] HINT:  Future log output will appear in directory "log".
. stopped waiting
pg_ctl: could not start server
Examine the log output.

启动pg失败,查看日志记录

2022-05-29 21:52:22.926 CST [35270] LOG:  starting PostgreSQL 14.3 on x86_64-pc-linux-gnu, 
            compiled by gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-44), 64-bit
2022-05-29 21:52:22.927 CST [35270] LOG:  listening on IPv6 address "::1", port 5432
2022-05-29 21:52:22.927 CST [35270] LOG:  listening on IPv4 address "127.0.0.1", port 5432
2022-05-29 21:52:22.929 CST [35270] LOG:  listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
2022-05-29 21:52:22.931 CST [35270] LOG:  listening on Unix socket "/tmp/.s.PGSQL.5432"
2022-05-29 21:52:22.936 CST [35272] LOG:  database system was interrupted; last known up at 2022-05-28 21:36:11 CST
2022-05-29 21:52:23.049 CST [35272] LOG:  invalid primary checkpoint record
2022-05-29 21:52:23.049 CST [35272] PANIC:  could not locate a valid checkpoint record
2022-05-29 21:52:24.211 CST [35270] LOG:  startup process (PID 35272) was terminated by signal 6: Aborted
2022-05-29 21:52:24.211 CST [35270] LOG:  aborting startup due to startup process failure
2022-05-29 21:52:24.218 CST [35270] LOG:  database system is shut down

错误比较明显,无法定位到有效的checkpoint记录,在oracle里面的意思可以理解为无法进行实例恢复,pg启动失败

重设wal
由于数据库为不一致状态,需要使用-f进行强制重设

-bash-4.2$ pg_resetwal $PGDATA
The database server was not shut down cleanly.
Resetting the write-ahead log might cause data to be lost.
If you want to proceed anyway, use -f to force reset.
-bash-4.2$ pg_resetwal -f $PGDATA
Write-ahead log reset

启动PG成功

-bash-4.2$ pg_ctl start -D $PGDATA
waiting for server to start....2022-05-29 22:01:02.647 CST [37178] LOG:  
redirecting log output to logging collector process
2022-05-29 22:01:02.647 CST [37178] HINT:  Future log output will appear in directory "log".
 done
server started

日志记录

2022-05-29 22:01:02.647 CST [37178] LOG:  starting PostgreSQL 14.3 on x86_64-pc-linux-gnu, 
                    compiled by gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-44), 64-bit
2022-05-29 22:01:02.648 CST [37178] LOG:  listening on IPv6 address "::1", port 5432
2022-05-29 22:01:02.648 CST [37178] LOG:  listening on IPv4 address "127.0.0.1", port 5432
2022-05-29 22:01:02.649 CST [37178] LOG:  listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
2022-05-29 22:01:02.651 CST [37178] LOG:  listening on Unix socket "/tmp/.s.PGSQL.5432"
2022-05-29 22:01:02.653 CST [37180] LOG:  database system was shut down at 2022-05-29 22:00:47 CST
2022-05-29 22:01:02.661 CST [37178] LOG:  database system is ready to accept connections

查看wal日志,产生新记录

-bash-4.2$ pwd
/var/lib/pgsql/14/data/pg_wal
-bash-4.2$ ls -ltr
total 16384
drwx------. 2 postgres postgres        6 May 24 02:20 archive_status
-rw-------. 1 postgres postgres 16777216 May 29 22:01 000000010000000000000014

验证刚刚创建测试表

-bash-4.2$ psql
psql (14.3)
Type "help" for help.

postgres=# select count(1) from t_xifenfei;
ERROR:  relation "t_xifenfei" does not exist
LINE 1: select count(1) from t_xifenfei;
                             ^

由于需要进行实例恢复的wal日志丢失导致这表记录也丢失.由此可见这类操作可能导致数据丢失风险,对于生产环境,需要慎重,

发表在 PostgreSQL, PostgreSQL恢复 | 标签为 , , | 评论关闭