postgres主从基于流复制

环境： CentOS Linux release 7.6.1810 (Core) 内核版本:3.10.0-957.10.1.el7.x86_64

node1:192.168.216.130

node2:192.168.216.132

node3:192.168.216.134

一、首先在3个节点分别安装postgres,这里由于线上环境使用9.5.3,故本人直接使用9.5.3用于实验测试,其他版本的编译安装方式是相同的

yum install gcc readline-devel zlib-devel
cd /tmp/
wget https://ftp.postgresql.org/pub/source/v9.5.3/postgresql-9.5.3.tar.gz
tar -xf postgresql-9.5.3.tar.gz
useradd postgres
./configure --prefix=/usr/local/postgresql
make -j2
make install
chown -R postgres:postgres /usr/local/postgresql/

二、 在2个节点上分别配置环境变量并执行初始化操作

1、切换用户
su - postgres
2、配置环境变量
vi .bash_profile

[postgres@localhost ~]$ cat .bash_profile
# .bash_profile

# Get the aliases and functions
if [ -f ~/.bashrc ]; then
	. ~/.bashrc
fi

# User specific environment and startup programs
PGHOME=/usr/local/postgresql
export PGHOME
PGDATA=/usr/local/postgresql/data
export PGDATA
PATH=$PATH:$HOME/.local/bin:$HOME/bin:$PGHOME/bin

export PATH
3、使上述环境立即变量生效
source .bash_profile
4、验证环境变量是否配置成功(查询postgres数据库版本号)
[postgres@localhost ~]$ psql -V
psql (PostgreSQL) 9.5.3
4、执行初始化：
initdb

三、编辑node1节点上的postgresql.conf配置文件

仅供参考，以下参数请根据需求进行合理配置修改
listen_addresses = '*'
port = 5432
wal_level = logical
archive_mode = on
max_wal_senders = 10
hot_standby = on
log_destination = 'csvlog'
logging_collector = on
log_directory = 'pg_log'
log_filename = 'postgresql-%Y-%m-%d_%H%M%S.log'
log_rotation_age = 1d
log_rotation_size = 20MB

四、编辑node1节点上的pg_hba.conf配置文件

# TYPE  DATABASE        USER            ADDRESS                 METHOD

# "local" is for Unix domain socket connections only
local   all             postgres                         trust
host    all             all             0.0.0.0/0        md5
# IPv4 local connections:
host    all             postgres        127.0.0.1/32            trust
# IPv6 local connections:
host    all             postgres             ::1/128            trust
# Allow replication connections from localhost, by a user with the
# replication privilege.
local   replication     replicator                              trust
host    replication     replicator     0.0.0.0/0              md5
host    replication     postgres        127.0.0.1/32            trust
#host    replication     postgres        ::1/128                trust

五、启动node1节点的数据库服务,并连接主库

pg_ctl -D /usr/local/postgresql/data start

验证主库是否正常
执行创建一个测试库sql语句，用于测试主库
create database test;
下图是我已经执行过的

执行sql语句创建基于流复制的用户，后面会用到，这里所创建的replicator用户名注意和上文的pg_hba.conf配置文件中的用户保持一致
create user replicator replication login encrypted password '1qaz2wsx';

六、创建物理复制槽，分别对应node2，node3

select * from pg_create_physical_replication_slot('pgsql95_132');
select * from pg_create_physical_replication_slot('pgsql95_134');

七、在node2、node3节点上清空初始化后的data数据

进入初始化目录
cd /usr/local/postgresql/data/
清空当前目录下的所有文件
rm -rf *

　八、在node2、node3节点上执行基础备份，配置stream replication

./pg_basebackup -h 192.168.216.130 -D /usr/local/postgresql/data -U replicator -c fast -X stream -v -P -R

这里的密码为创建replicator用户时所指定的密码

编辑node2、node3上的recovery.conf文件，注意 recovery.conf 的 primary_slot_name 在不同节点值会不同。

node2：

[postgres@localhost data]$ cat recovery.conf
standby_mode = 'on'
recovery_target_timeline = 'latest'
primary_conninfo = 'user=replicator password=1qaz2wsx host=192.168.216.130 port=5432 sslmode=disable sslcompression=1'
primary_slot_name = 'pgsql95_132'

node3：

[postgres@localhost data]$ cat recovery.conf
standby_mode = 'on'
recovery_target_timeline = 'latest'
primary_conninfo = 'user=replicator password=1qaz2wsx host=192.168.216.130 port=5432 sslmode=disable sslcompression=1'
primary_slot_name = 'pgsql95_134'

九、分别启动node2、node3节点的postgres服务

pg_ctl -D /usr/local/postgresql/data start

　此时在node1节点上执行以下sql语句，可以看到active字段由原来的“f”变为“t”

且在主库上(node1)可以看到postgres进程为wal sender状态

两个从库上node2，node3上的postgres进程为wal receiver状态

查询复制状态

select client_addr,
       pg_xlog_location_diff(sent_location, write_location) as write_delay,
       pg_xlog_location_diff(sent_location, flush_location) as flush_delay,
       pg_xlog_location_diff(sent_location, replay_location) as replay_delay
 from pg_stat_replication;