PostgreSQL的一主一从集群搭建部署 (同步)

keep__go2024-12-11 8:07

一、实验环境

虚拟机名 IP 身份 简称
keep-postgres12-node1 192.168.122.87 主节点 node1
keep-postgres12-node2 192.168.122.89 备节点 node2

二、安装数据库

源码包方式（主）
1、创建用户
shell 复制代码 
[root@keep-postgres12-node1 ~]# groupadd postgres
[root@keep-postgres12-node1 ~]# useradd -g postgres postgres -m -s /bin/bash
[root@keep-postgres12-node1 ~]# echo "Database@123" | passwd --stdin postgres
2、修改服务器内核信息
shell 复制代码 
[root@keep-postgres12-node1 ~]# echo '# 最大共享内存段大小 (默认值68719476736)
kernel.shmmax = 68719476736 
# 可以使用的共享内存的总量 (默认值4294967296)
kernel.shmall = 4294967296 
# 整个系统共享内存段的最大数目
kernel.shmmni = 4096 
# 每个信号对象集的最大信号对象数
kernel.sem = 50100 64128000 50100 1280 
# 文件句柄的最大数量
fs.file-max = 7672460 
# 应用程序可使用的IPv4端口范围
net.ipv4.ip_local_port_range = 9000 65000
# 套接字接收缓冲区大小的缺省值
net.core.rmem_default = 1048576 
# 套接字发送缓冲区大小的缺省值
net.core.wmem_default = 262144 
# 套接字发送缓冲区大小的最大值
net.core.wmem_max = 1048576' >> /etc/sysctl.conf

# 使参数生效
[root@keep-postgres12-node1 ~]# sysctl -p
3、安装依赖包
shell 复制代码 
[root@keep-postgres12-node1 ~]# yum install gcc gcc-c++ zlib-devel readline-devel perl-ExtUtils-Embed pam-devel openssl openssl-devel cmake libxslt-devel libxml2-devel openldap-devel python-devel tcl tcl-devel bison flex xmlto -y
4、创建数据库目录
shell 复制代码 
[root@keep-postgres12-node1 ~]# mkdir /data
[root@keep-postgres12-node1 ~]# mkdir /data/postgres12.2
[root@keep-postgres12-node1 ~]# chown -R postgres: /data
5、创建环境变量
shell 复制代码 
[root@keep-postgres12-node1 ~]# su - postgres
[postgres@keep-postgres12-node1 ~]$ echo 'export PGPORT=5432
export PG_HOME=/data/postgres12.2
export PATH=$PG_HOME/bin:$PATH
export PGDATA=$PG_HOME/data
export LD_LIBRARY_PATH=$PG_HOME/lib
export LANG=en_US.utf8' >> ~/.bash_profile && source  ~/.bash_profile
6、下载源码包，并解压

链接1：https://ftp.postgresql.org/pub/source

链接2：https://www.postgresql.org/ftp/source/

shell 复制代码 
# 安装postgresql-12.2版本
# 下载postgresql-12.2源码包
[postgres@keep-postgres12-node1 ~]$ cd /data
[postgres@keep-postgres12-node1 data]$ wget https://ftp.postgresql.org/pub/source/v12.2/postgresql-12.2.tar.bz2
shell 复制代码 
# 解压源码包
[postgres@keep-postgres12-node1 data]$ tar -xf postgresql-12.2.tar.bz2
7、编译源码包
shell 复制代码 
[postgres@keep-postgres12-node1 postgresql-12.2]$ ./configure --prefix=/data/postgres12.2 --with-pgport=5432 --with-openssl --with-perl \
--with-tcl --with-python --with-pam --without-ldap --with-libxml --with-libxslt \
--enable-thread-safety --with-wal-blocksize=16 --with-blocksize=8
shell 复制代码 
# 包括第三方插件全部编译，包含文档和所有的contirb
[postgres@keep-postgres12-node1 postgresql-12.2]$ gmake world
8、安装数据库
shell 复制代码 
[postgres@keep-postgres12-node1 postgresql-12.2]$ make && make install
9、配置wal日志目录
shell 复制代码 
[postgres@keep-postgres12-node1 postgresql-12.2]$ # 用于保存wal日志
[postgres@keep-postgres12-node1 postgresql-12.2]$ mkdir $PG_HOME/arch
10、初始化数据库
shell 复制代码 
[postgres@keep-postgres12-node1 postgresql-12.2]$ # 创建目录
[postgres@keep-postgres12-node1 postgresql-12.2]$ mkdir $PG_HOME/data
[postgres@keep-postgres12-node1 postgresql-12.2]$ # 初始化数据库集簇
[postgres@keep-postgres12-node1 postgresql-12.2]$ # --data-checksums 主从复制时需要
[postgres@keep-postgres12-node1 postgresql-12.2]$ initdb -D $PGDATA -W --data-checksums
11、启动数据库
shell 复制代码 
[postgres@keep-postgres12-node1 postgresql-12.2]$ pg_ctl -D $PGDATA start
12、创建数据库用户和数据库
shell 复制代码 
[postgres@keep-postgres12-node1 data]$ psql -d postgres -p 5432
postgres=# create database testdb;
postgres=# create user keep with Superuser password  'keep';
13、配置远程连接

# 配置连接

shell 复制代码 
[postgres@keep-postgres12-node1 data]$ cd $PGDATA
[postgres@keep-postgres12-node1 data]$ sed -i "s/#listen_addresses = 'localhost'/listen_addresses = '*'/g" postgresql.conf
[postgres@keep-postgres12-node1 data]$ echo 'host all all 0.0.0.0/0 md5' >> pg_hba.conf

# 重启数据库

shell 复制代码 
[postgres@keep-postgres12-node1 data]$ pg_ctl stop
[postgres@keep-postgres12-node1 data]$ pg_ctl -D $PGDATA start

# 验证远程连接

shell 复制代码 
[postgres@keep-postgres12-node1 data]$ psql -d testdb -U keep -p 5432 -h 192.168.140.96
14、配置数据库日志
shell 复制代码 
[postgres@keep-postgres12-node1 data]$ echo "# Add settings for extensions here
log_destination='csvlog'
logging_collector=on
log_directory='pg_log'
log_filename='postgresql-%Y-%m-%d.log'
log_truncate_on_rotation=off
log_rotation_age=1d
log_rotation_size=0
log_error_verbosity=verbose" >> $PGDATA/postgresql.conf
源码包方式（备）
1、创建用户
shell 复制代码 
[root@keep-postgres12-node2 ~]# groupadd postgres
[root@keep-postgres12-node2 ~]# useradd -g postgres postgres -m -s /bin/bash
[root@keep-postgres12-node2 ~]# echo "Database@123" | passwd --stdin postgres
2、修改服务器内核信息
shell 复制代码 
[root@keep-postgres12-node2 ~]# echo '# 最大共享内存段大小 (默认值68719476736)
kernel.shmmax = 68719476736 
# 可以使用的共享内存的总量 (默认值4294967296)
kernel.shmall = 4294967296 
# 整个系统共享内存段的最大数目
kernel.shmmni = 4096 
# 每个信号对象集的最大信号对象数
kernel.sem = 50100 64128000 50100 1280 
# 文件句柄的最大数量
fs.file-max = 7672460 
# 应用程序可使用的IPv4端口范围
net.ipv4.ip_local_port_range = 9000 65000
# 套接字接收缓冲区大小的缺省值
net.core.rmem_default = 1048576 
# 套接字发送缓冲区大小的缺省值
net.core.wmem_default = 262144 
# 套接字发送缓冲区大小的最大值
net.core.wmem_max = 1048576' >> /etc/sysctl.conf

# 使参数生效
[root@keep-postgres12-node2 ~]# sysctl -p
3、安装依赖包
shell 复制代码 
[root@keep-postgres12-node2 ~]# yum install gcc gcc-c++ zlib-devel readline-devel perl-ExtUtils-Embed pam-devel openssl openssl-devel cmake libxslt-devel libxml2-devel openldap-devel python-devel tcl tcl-devel bison flex xmlto -y
4、创建数据库目录
shell 复制代码 
[root@keep-postgres12-node2 ~]# mkdir /data
[root@keep-postgres12-node2 ~]# mkdir /data/postgres12.2
[root@keep-postgres12-node2 ~]# chown -R postgres: /data
5、创建环境变量
shell 复制代码 
[root@keep-postgres12-node2 ~]# su - postgres
[postgres@keep-postgres12-node2 ~]$ echo 'export PGPORT=5432
export PG_HOME=/data/postgres12.2
export PATH=$PG_HOME/bin:$PATH
export PGDATA=$PG_HOME/data
export LD_LIBRARY_PATH=$PG_HOME/lib
export LANG=en_US.utf8' >> ~/.bash_profile && source  ~/.bash_profile
6、下载源码包，并解压

链接1：https://ftp.postgresql.org/pub/source

链接2：https://www.postgresql.org/ftp/source/

shell 复制代码 
# 安装postgresql-12.2版本
# 下载postgresql-12.2源码包
[postgres@keep-postgres12-node2 ~]$ cd /data
[postgres@keep-postgres12-node2 data]$ wget https://ftp.postgresql.org/pub/source/v12.2/postgresql-12.2.tar.bz2
shell 复制代码 
# 解压源码包
[postgres@keep-postgres12-node2 data]$ tar -xf postgresql-12.2.tar.bz2
7、编译源码包
shell 复制代码 
[postgres@keep-postgres12-node2 postgresql-12.2]$ ./configure --prefix=/data/postgres12.2 --with-pgport=5432 --with-openssl --with-perl \
--with-tcl --with-python --with-pam --without-ldap --with-libxml --with-libxslt \
--enable-thread-safety --with-wal-blocksize=16 --with-blocksize=8
shell 复制代码 
# 包括第三方插件全部编译，包含文档和所有的contirb
[postgres@keep-postgres12-node2 postgresql-12.2]$ gmake world
8、安装数据库
shell 复制代码 
[postgres@keep-postgres12-node2 postgresql-12.2]$ make && make install
9、配置wal日志目录
shell 复制代码 
[postgres@keep-postgres12-node1 postgresql-12.2]$ # 用于保存wal日志
[postgres@keep-postgres12-node1 postgresql-12.2]$ mkdir $PG_HOME/arch
配置互信
1、配置主机名
shell 复制代码 
# keep-postgres12-node1
[root@keep-postgres12-node1 ~]# echo '192.168.122.87 keep-postgres12-node1
192.168.122.88 keep-postgres12-node2' >> /etc/hosts

# keep-postgres12-node2
[root@keep-postgres12-node2 ~]# echo '192.168.122.87 keep-postgres12-node1
192.168.122.88 keep-postgres12-node2' >> /etc/hosts
2、配置postgres用户互信

# keep-postgres12-node1

shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ ssh-keygen
[postgres@keep-postgres12-node1 ~]$ ssh-copy-id keep-postgres12-node1
[postgres@keep-postgres12-node1 ~]$ ssh-copy-id keep-postgres12-node2

# 验证
[postgres@keep-postgres12-node1 ~]$ ssh keep-postgres12-node1
[postgres@keep-postgres12-node1 ~]$ ssh keep-postgres12-node2

# keep-postgres12-node2

shell 复制代码 
[postgres@keep-postgres12-node2 ~]$ ssh-keygen
[postgres@keep-postgres12-node2 ~]$ ssh-copy-id keep-postgres12-node1
[postgres@keep-postgres12-node2 ~]$ ssh-copy-id keep-postgres12-node2

# 验证
[postgres@keep-postgres12-node2 ~]$ ssh keep-postgres12-node1
[postgres@keep-postgres12-node2 ~]$ ssh keep-postgres12-node2

三、修改主节点配置

1、创建同步用户
shell 复制代码 
# 数据库需要在线状态
[postgres@keep-postgres12-node1 ~]$ psql -c "create role replrole login replication encrypted password 'ReplRole@123';"
2、配置$PGDATA/pg_hba.conf
shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ echo 'host  replication  replrole  keep-postgres12-node1  trust
host  replication  replrole  keep-postgres12-node2  trust
' >> $PGDATA/pg_hba.conf
3、修改数据库参数
shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ echo "wal_level = replica
max_wal_senders=20
wal_keep_segments =64
# 开启归档
archive_mode = on
archive_command = 'cp %p /data/postgres12.2/arch/%f'
restore_command = 'cp /data/postgres12.2/arch/%f %p'
recovery_target_timeline = 'latest'
# full_page_writes是控制是否开启全页写入
full_page_writes = on
# 将每个磁盘页的全部内容写入到WAL
wal_log_hints = on 
synchronous_standby_names = 'standby_pg2' 
# 默认值，可以设置为remote_write，对主库性能有利
# 开启同步模式，此刻需要强同步，数据无法入库
synchronous_commit = on 
" >> $PGDATA/postgresql.conf
4、重启数据库
shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ pg_ctl restart

四、修改备节点配置

1、进行数据恢复
shell 复制代码 
[postgres@keep-postgres12-node2 ~]$ pg_basebackup -h keep-postgres12-node1 -p 5432 -U replrole -R -F p -P -D $PG_HOME/data
2、修改standby.signal配置
shell 复制代码 
[postgres@keep-postgres12-node2 data]$ cd $PGDATA
[postgres@keep-postgres12-node2 data]$ cat standby.signal
primary_conninfo = 'host=keep-postgres12-node1 port=5432 user=replrole password=ReplRole@123 application_name=standby_pg2 options=''-c wal_sender_timeout=5000'''
restore_command = 'cp /data/postgres12.2/arch/%f %p'
archive_cleanup_command = 'pg_archivecleanup /data/postgres12.2/arch %r'
standby_mode = on
3、修改postgresql.auto.conf配置

# 注意：强同步需要注意在postgresql.auto.conf中加入application_name=standby_pg2

shell 复制代码 
[postgres@keep-postgres12-node2 data]$ cat postgresql.auto.conf
primary_conninfo = 'user=replrole passfile=''/home/postgres/.pgpass'' host=''keep-postgres12-node1'' application_name=standby_pg2 port=5432 sslmode=prefer sslcompression=0 gssencmode=disable krbsrvname=postgres target_session_attrs=any'

五、启动数据同步

# 注意：主库修改synchronous_standby_names = 'standby_pg2' 后，需等待备库接收主库发送的WAL日志流并写入WAL文件，之后才向客户端返回成功。

1、主库检查点
shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ pg_controldata
pg_control version number:            1201
Catalog version number:               201909212
Database system identifier:           7435168474920575893
Database cluster state:               in production
pg_control last modified:             Mon 02 Dec 2024 11:14:28 AM CST
Latest checkpoint location:           0/1F000028
Latest checkpoint's REDO location:    0/1F000028
Latest checkpoint's REDO WAL file:    00000003000000000000001F
Latest checkpoint's TimeLineID:       3
Latest checkpoint's PrevTimeLineID:   3
Latest checkpoint's full_page_writes: on
Latest checkpoint's NextXID:          0:609
Latest checkpoint's NextOID:          16643
Latest checkpoint's NextMultiXactId:  1
Latest checkpoint's NextMultiOffset:  0
Latest checkpoint's oldestXID:        480
Latest checkpoint's oldestXID's DB:   1
Latest checkpoint's oldestActiveXID:  0
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:            Mon 02 Dec 2024 11:14:27 AM CST
Fake LSN counter for unlogged rels:   0/3E8
Minimum recovery ending location:     0/0
Min recovery ending loc's timeline:   0
Backup start location:                0/0
Backup end location:                  0/0
End-of-backup record required:        no
wal_level setting:                    replica
wal_log_hints setting:                on
max_connections setting:              100
max_worker_processes setting:         8
max_wal_senders setting:              20
max_prepared_xacts setting:           0
max_locks_per_xact setting:           64
track_commit_timestamp setting:       off
Maximum data alignment:               8
Database block size:                  8192
Blocks per segment of large relation: 131072
WAL block size:                       16384
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
Float4 argument passing:              by value
Float8 argument passing:              by value
Data page checksum version:           1
Mock authentication nonce:            2fd610c9c82cf604c47448c0e7c842aa9ac4944e8fea828fcc0d2425b6f3a6e4
2、启动备库
shell 复制代码 
[postgres@keep-postgres12-node2 data]$ pg_ctl start
3、查看备库日志
shell 复制代码 
[postgres@keep-postgres12-node2 pg_log]$ tail -f postgresql-2024-12-02.csv

2024-12-02 11:19:12.451 CST,,,25290,,674d2730.62ca,2,,2024-12-02 11:19:12 CST,,0,LOG,00000,"entering standby mode",,,,,,,,"StartupXLOG, xlog.c:6324",""
2024-12-02 11:19:12.462 CST,,,25290,,674d2730.62ca,3,,2024-12-02 11:19:12 CST,1/0,0,LOG,00000,"redo starts at 0/1F0000A0",,,,,,,,"StartupXLOG, xlog.c:7037",""
2024-12-02 11:19:12.462 CST,,,25290,,674d2730.62ca,4,,2024-12-02 11:19:12 CST,1/0,0,LOG,00000,`"consistent recovery state reached at 0/1F000350",,,,,,,,"CheckRecoveryConsistency, xlog.c:7880",""
2024-12-02 11:19:12.462 CST,,,25290,,674d2730.62ca,5,,2024-12-02 11:19:12 CST,1/0,0,LOG,00000,"invalid record length at 0/1F000350: wanted 24, got 0",,,,,,,,"ReadRecord, xlog.c:4284",""
2024-12-02 11:19:12.462 CST,,,25288,,674d2730.62c8,2,,2024-12-02 11:19:12 CST,,0,LOG,00000,"database system is ready to accept read only connections",,,,,,,,"sigusr1_handler, postmaster.c:5153",""
2024-12-02 11:19:12.473 CST,,,25297,,674d2730.62d1,1,,2024-12-02 11:19:12 CST,,0,LOG,00000,"started streaming WAL from primary at 0/1F000000 on timeline 3",,,,,,,,"WalReceiverMain, walreceiver.c:371",""
4、检查数据库进程
shell 复制代码 
# keep-postgres12-node1
[postgres@keep-postgres12-node1 ~]$ ps -ef | grep postgres:
postgres 28608 28601  0 11:14 ?        00:00:00 postgres: logger
postgres 28610 28601  0 11:14 ?        00:00:00 postgres: checkpointer
postgres 28611 28601  0 11:14 ?        00:00:00 postgres: background writer
postgres 28612 28601  0 11:14 ?        00:00:00 postgres: walwriter
postgres 28613 28601  0 11:14 ?        00:00:00 postgres: autovacuum launcher
postgres 28614 28601  0 11:14 ?        00:00:00 postgres: archiver
postgres 28615 28601  0 11:14 ?        00:00:00 postgres: stats collector
postgres 28616 28601  0 11:14 ?        00:00:00 postgres: logical replication launcher
postgres 29485 28601  0 11:18 ?        00:00:00 postgres: postgres testdb [local] idle
postgres 29845 28601  0 11:19 ?        00:00:00 postgres: `walsender replrole 192.168.122.89(54086) streaming 0/1F000438`


# keep-postgres12-node2
[postgres@keep-postgres12-node2 data]$ ps -ef | grep postgres:
postgres 25289 25288  0 11:19 ?        00:00:00 postgres: logger
postgres 25290 25288  0 11:19 ?        00:00:00 postgres: startup   recovering 00000003000000000000001F
postgres 25294 25288  0 11:19 ?        00:00:00 postgres: checkpointer
postgres 25295 25288  0 11:19 ?        00:00:00 postgres: background writer
postgres 25296 25288  0 11:19 ?        00:00:00 postgres: stats collector
postgres 25297 25288  0 11:19 ?        00:00:00 postgres: `walreceiver   streaming 0/1F000438`

六、验证数据同步

1、主库检查同步状态sync_state 为sync
shell 复制代码 
[postgres@keep-postgres12-node1 data]$ psql
postgres=# \x
Expanded display is on.
postgres=#  SELECT * FROM pg_stat_replication;
-[ RECORD 1 ]----+------------------------------
pid              | 29845
usesysid         | 16642
usename          | replrole
application_name | standby_pg2
client_addr      | 192.168.122.89
client_hostname  | keep-postgres12-node2
client_port      | 54086
backend_start    | 2024-12-02 11:19:12.468565+08
backend_xmin     |
state            | streaming
sent_lsn         | 0/1F000438
write_lsn        | 0/1F000438
flush_lsn        | 0/1F000438
replay_lsn       | 0/1F000438
write_lag        |
flush_lag        |
replay_lag       |
sync_priority    | 1
sync_state       | sync
reply_time       | 2024-12-02 11:23:49.82386+08
2、测试数据同步
shell 复制代码 
-- 主库执行
drop table if exists employees_sync CASCADE;
CREATE TABLE employees (
    id SERIAL PRIMARY KEY,
    name VARCHAR(100),
    position VARCHAR(100),
    department VARCHAR(100),
    hire_date DATE
);

INSERT INTO employees_sync (name, position, department, hire_date)
SELECT 
    'Employee ' || (generate_series(1, 200))::text,
    'Position ' || (generate_series(1, 200))::text,
    'Department ' || (random() *(200-1)+1)::text,
    '2010-01-01'::date + (generate_series(1, 200) * interval '1 day');

-- 备库查询数据
select count(1) from employees_sync ;
-- 计算两边的md5值
SELECT md5(
	string_agg(
    id::text || '-' ||
    name || '-' ||
    position || '-' ||
    department || '-' ||
    TO_CHAR(hire_date, 'YYYY-MM-DD'),',')
) AS row_md5
FROM employees_sync;

七、主备切换

1、停主库
shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ pg_ctl stop -m fast
2、主从切换
备库升主
shell 复制代码 
# 切换之后，$PGDATA下原有的 standby.signal 文件不存在了
[postgres@keep-postgres12-node2 ~]$ pg_ctl promote
查看状态
shell 复制代码 
[postgres@keep-postgres12-node2 ~]$ pg_controldata | grep -i cluster
Database cluster state:               `in production`
3、检查原备库的配置（现主）
检查配置文件postgresql.auto.conf
shell 复制代码 
[postgres@keep-postgres12-node2 data]$ cat postgresql.auto.conf
# Do not edit this file manually!
# It will be overwritten by the ALTER SYSTEM command.
# 检查是否进行注释，配置还是当作备库，自相矛盾
# primary_conninfo = 'user=replrole passfile=''/home/postgres/.pgpass'' host=''keep-postgres12-node1'' application_name=standby_pg2 port=5432 sslmode=prefer sslcompression=0 gssencmode=disable krbsrvname=postgres target_session_attrs=any'
检查状态
shell 复制代码 
# 此处进程应为 walwriter，表示还是备机状态。若修改了postgresql.auto.conf，需重启数据库
[postgres@keep-postgres12-node2 data]$ ps -ef | grep postgres
postgres 25288     1  0 Dec02 ?        00:00:00 /data/postgres12.2/bin/postgres
postgres 25289 25288  0 Dec02 ?        00:00:00 postgres: logger
postgres 25294 25288  0 Dec02 ?        00:00:02 postgres: checkpointer
postgres 25295 25288  0 Dec02 ?        00:00:02 postgres: background writer
postgres 25296 25288  0 Dec02 ?        00:00:00 postgres: stats collector
postgres 29058 25288  0 16:11 ?        00:00:00 postgres: `walwriter`
postgres 29059 25288  0 16:11 ?        00:00:00 postgres: autovacuum launcher
postgres 29060 25288  0 16:11 ?        00:00:00 postgres: archiver   last was 000000030000000000000020.partial
postgres 29061 25288  0 16:11 ?        00:00:00 postgres: logical replication launcher

[postgres@keep-postgres12-node2 data]$ pg_ctl restart

# 重启数据库后，备机状态为未连接，不会启动walsender
[postgres@keep-postgres12-node2 data]$ ps -ef | grep postgres
postgres 32261     1  0 16:20 ?        00:00:00 /data/postgres12.2/bin/postgres
postgres 32262 32261  0 16:20 ?        00:00:00 postgres: logger
postgres 32264 32261  0 16:20 ?        00:00:00 postgres: checkpointer
postgres 32265 32261  0 16:20 ?        00:00:00 postgres: background writer
postgres 32266 32261  0 16:20 ?        00:00:00 postgres: walwriter
postgres 32267 32261  0 16:20 ?        00:00:00 postgres: autovacuum launcher
postgres 32268 32261  0 16:20 ?        00:00:00 postgres: archiver
postgres 32269 32261  0 16:20 ?        00:00:00 postgres: stats collector
postgres 32270 32261  0 16:20 ?        00:00:00 postgres: logical replication launcher
4、备库配置standby.signal文件
shell 复制代码 
# standby.signal文件需要新建
[postgres@keep-postgres12-node1 data]$ cat standby.signal
# 添加以下内容
primary_conninfo = 'host=keep-postgres12-node2 port=5432 user=replrole password=ReplRole@123 application_name=standby_pg2 options=''-c wal_sender_timeout=5000'''
restore_command = 'cp /data/postgres12.2/arch/%f %p'
archive_cleanup_command = 'pg_archivecleanup /data/postgres12.2/arch %r'
standby_mode = on
5、备库修改配置文件postgresql.auto.conf
shell 复制代码 
[postgres@keep-postgres12-node1 data]$ cat postgresql.auto.conf
# 添加以下内容
primary_conninfo = 'user=replrole passfile=''/home/postgres/.pgpass'' host=''keep-postgres12-node2'' application_name=standby_pg2 port=5432 sslmode=prefer sslcompression=0 gssencmode=disable krbsrvname=postgres target_session_attrs=any'
6、启动备库
shell 复制代码 
[postgres@keep-postgres12-node1 data]$ pg_ctl start
7、检查数据库进程
shell 复制代码 
# 备库
[postgres@keep-postgres12-node1 ~]$ ps -ef | grep postgres
postgres 32638     1  0 16:38 ?        00:00:00 /data/postgres12.2/bin/postgres
postgres 32639 32638  0 16:38 ?        00:00:00 postgres: logger
postgres 32640 32638  0 16:38 ?        00:00:00 postgres: startup   recovering 000000040000000000000021
postgres 32644 32638  0 16:38 ?        00:00:00 postgres: checkpointer
postgres 32645 32638  0 16:38 ?        00:00:00 postgres: background writer
postgres 32646 32638  0 16:38 ?        00:00:00 postgres: stats collector
postgres 32647 32638  0 16:38 ?        00:00:00 postgres: `walreceiver`   streaming 0/21002210

# 主库
[postgres@keep-postgres12-node2 ~]$ ps -ef | grep postgres
postgres  4764 32261  0 16:38 ?        00:00:00 postgres: `walsender` replrole 192.168.122.87(18913) streaming 0/21002210
postgres 32261     1  0 16:20 ?        00:00:00 /data/postgres12.2/bin/postgres
postgres 32262 32261  0 16:20 ?        00:00:00 postgres: logger
postgres 32264 32261  0 16:20 ?        00:00:00 postgres: checkpointer
postgres 32265 32261  0 16:20 ?        00:00:00 postgres: background writer
postgres 32266 32261  0 16:20 ?        00:00:00 postgres: walwriter
postgres 32267 32261  0 16:20 ?        00:00:00 postgres: autovacuum launcher
postgres 32268 32261  0 16:20 ?        00:00:00 postgres: archiver
postgres 32269 32261  0 16:20 ?        00:00:00 postgres: stats collector
postgres 32270 32261  0 16:20 ?        00:00:00 postgres: logical replication launcher
8、验证是否同步
sql 复制代码 
-- 主库执行
drop table if exists employees1_sync CASCADE;
CREATE TABLE employees1_sync (
    id SERIAL PRIMARY KEY,
    name VARCHAR(100),
    position VARCHAR(100),
    department VARCHAR(100),
    hire_date DATE
);

INSERT INTO employees1_sync (name, position, department, hire_date)
SELECT 
    'Employee ' || (generate_series(1, 200))::text,
    'Position ' || (generate_series(1, 200))::text,
    'Department ' || (random() *(200-1)+1)::text,
    '2010-01-01'::date + (generate_series(1, 200) * interval '1 day');

-- 备库查询数据
select count(1) from employees1_sync ;
-- 计算两边的md5值
SELECT md5(
	string_agg(
    id::text || '-' ||
    name || '-' ||
    position || '-' ||
    department || '-' ||
    TO_CHAR(hire_date, 'YYYY-MM-DD'),',')
) AS row_md5
FROM employees1_sync;

八、pg_rewind 工具

​ 当主备集群中的备库意外崩溃，且经过长时间，归档日志又被删除了，需要把这段时间的增量数据同步回来，那么就可以用 到pg_rewind 工具进行同步。pg_rewind 使一个PostgreSQL 数据目录与另一个数据目录一致。

shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ pg_rewind --target-pgdata $PGDATA --source-server='host=keep-postgres12-node2 port=5432 user=keep password=keep dbname=postgres' -P
pg_rewind帮助命令
shell 复制代码 
[postgres@keep-postgres12-node1 ~]$ pg_rewind --help
pg_rewind resynchronizes a PostgreSQL cluster with another copy of the cluster.

Usage:
  pg_rewind [OPTION]...

Options:
  -D, --target-pgdata=DIRECTORY  existing data directory to modify
      --source-pgdata=DIRECTORY  source data directory to synchronize with
      --source-server=CONNSTR    source server to synchronize with
  -n, --dry-run                  stop before modifying anything
  -N, --no-sync                  do not wait for changes to be written
                                 safely to disk
  -P, --progress                 write progress messages
      --debug                    write a lot of debug messages
  -V, --version                  output version information, then exit
  -?, --help                     show this help, then exit
相关推荐
python机器学习建模3 小时前
科研论文必须要了解的25个学术网站
数据库
J.P.August4 小时前
Oracle DataGuard启动与关闭顺序
数据库·oracle
尚雷55804 小时前
Oracle 与 达梦 数据库 对比
数据库·oracle·达梦数据库
小猿姐6 小时前
Ape-DTS：开源 DTS 工具，助力自建 MySQL、PostgreSQL 迁移上云
数据库·mysql·postgresql·开源
百香果果ccc6 小时前
MySQL中的单行函数和聚合函数
数据库·mysql
摸摸陌陌6 小时前
Redis快速入门
数据库·redis·缓存
Elastic 中国社区官方博客6 小时前
Elasticsearch Serverless 中的数据流自动分片
大数据·数据库·elasticsearch·搜索引擎·serverless·时序数据库
Minyy116 小时前
牛客网刷题SQL--高级查询
数据库·sql
秋意钟6 小时前
MySQL基本架构
数据库·mysql·架构
朱小勇本勇7 小时前
Qt实现控件拖曳
开发语言·数据库·qt
热门推荐
01Windows11 Copilot助手开启教程(免费GPT-4)0204 - matlab m_map地学绘图工具基础函数 - 设置网格03玄机平台应急响应—webshell查杀04『手写ahooks系列-useClickAway』一个hook解决监听点击元素外部05【halcon】halcon字符识别——OCR06(欧拉)openEuler系统添加网卡文件配置流程、(欧拉)openEuler系统手动配置ipv6地址流程、(欧拉)openEuler系统网络管理说明07量化方法怎么选？如何评估量化后的大模型LLM？08支付宝短视频平台创作分成激励项目09Docker 夺命连环 15 问10MacBook外接键盘修改键位