实验环境
| 名称 |
IP |
系统 |
| Master |
172.17.10.190 |
|
| Slave |
172.17.10.189 |
CentOS 6.5 |
1.yun安装
|
1
2
|
rpm -ivh https:
//download
.postgresql.org
/pub/repos/yum/9
.6
//rhel-6-x86_64/pgdg-centos96-9
.6-3.noarch.rpm
yum
install
postgresql96.x86_64 postgresql96-server.x86_64 -y
|
2.主从配置
2.1 主数据库配置
启动master
|
1
2
3
4
|
/etc/init
.d
/postgresql-9
.6 initdb
/etc/init
.d
/postgresql-9
.6 start
su
- postgres
psql
|
授权
|
1
|
create role repl login replication encrypted password
'51idc.com'
;
|
编辑hba文件
/var/lib/pgsql/9.6/data/pg_hba.conf
新增
|
1
2
|
host replication repl 172.17.10.0
/24
md5
host all repl 172.17.10.0
/24
trust
|
编辑配置文件
/var/lib/pgsql/9.6/data/postgresql.conf
|
1
2
3
4
5
6
7
8
|
listen_addresses = 172.17.10.190
wal_level = hot_standby
#热备模式
max_wal_senders= 6
#可以设置最多几个流复制链接，差不多有几个从，就设置多少
wal_keep_segments = 10240
#重要配置
wal_send_timeout = 60s
max_connections = 512
#从库的 max_connections要大于主库
archive_mode = on
#允许归档
archive_command =
'cp %p /url/path%f'
#根据实际情况设置
|
2.2 从数据库配置
|
1
|
su
- postgres
|
如果开始为启动数据库可忽略下一步
|
1
2
3
|
rm
-rf
/var/lib/pgsql/9
.6
/data/
*
#开始没有启动从数据库，这一步可以省略
pg_basebackup -h 172.17.10.190 -U repl -D
/var/lib/pgsql/9
.6
/data
-X stream -P
cp
/usr/pgsql-9
.6
/share/recovery
.conf.sample
/var/lib/pgsql/9
.6
/data/recovery
.conf
|
修改配置文件recovery.conf
|
1
2
3
4
|
standby_mode = on
primary_conninfo =
'host=172.17.10.190 port=5432 user=repl password=51idc.com'
trigger_file =
'/var/lib/pgsql/9.6/data/trigger.kenyon'
#主从切换时后的触发文件
recovery_target_timeline =
'latest'
|
配置postgresql.conf文件
|
1
2
3
4
5
6
7
|
listen_addresses = 172.17.10.189
wal_level = hot_standby
max_connections = 1000
#一般从的最大链接要大于主的。
hot_standby = on
#说明这台机器不仅仅用于数据归档，也用于查询
max_standby_streaming_delay = 30s
wal_receiver_status_interval = 10s
#多久向主报告一次从的状态。
hot_standby_feedback = on
#如果有错误的数据复制，是否向主进行范例
|
检测
|
1
|
select
client_addr,sync_state from pg_stat_replication;
|
查看主从状态
|
1
|
select
* from pg_stat_replication;
|
脚本监控主从
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
|
#!/bin/bash
# mail [email protected]
data=`
date
+%Y-%M-%d
" "
%H:%m`
netstat
-lntup|
grep
5432 &&
ps
-ef|
grep
postmaster
if
[ $? -
eq
0 ];
then
for
IP
in
172.17.10.188 172.17.10.189
do
/usr/bin/psql
-h 172.17.10.190 -p 5432 -U repl -d postgres --
command
"select * from pg_stat_replication"
|
grep
$IP
if
[
"$?"
!=
"0"
];
then
echo
"postgresql master-slave status is error! please login check!"
|mail -r
&&
echo
"$data postgresql postgresql master-slave status is error!"
>>
/var/log/postgresql-error
.log
fi
done
else
echo
"postgresql master-slave status is error! please login check!"
|mail -r
&&
echo
"$data postgresql postgresql master-slave status is error!"
>>
/var/log/postgresql-error
.log
fi
|
2.3主从切换
主库查看进程为sender
备库
停止主库
查看slave的日志
创建触发文件，切换主
|
1
|
touch
trigger.kenyon
|
查看slave的日志，面前已经切换为主
使用pg_controldata
备机状态为： in archive recovery
主库状态为：in production