Slony是PostgreSQL领域中最广泛的复制解决方案之一。它不仅是最古老的复制实现之一，它也是一个拥有最广泛的外部工具支持的工具，Slony使用逻辑复制；Slony-I一般要求表有主键,或者唯一键；Slony的工作不是基于PostgreSQL事务日志的，而是基于触发器的；基于逻辑复制高可用性，PostgreSQL除了slony；还有Londiste,BDR等等后续文章会讲到。

1.安装Slony

下载地址：http://www.slony.info
安装步骤

tar -jxvf slony1-2.2.5.tar.bz2
cd slony1-2.2.5
./configure --with-pgconfigdir=/opt/pgsql96/bin
make
make install

安装完成！
执行./configure时；会在当前目录是否可以找到pg_config命令；本例pg_config在/opt/pgsql96/bin目录下；

2. Slony架构图

slony.png

3. 复制表

现有实验环境

主机名	IP	角色
PostgreSQL201	192.168.1.201	master
PostgreSQL202	192.168.1.202	slave

3.1 在两台数据库中都创建一个slony的超级用户；专为slony服务；

create user slony superuser password 'li0924';

3.2 本实验两台主机都有lottu数据库；以lottu数据库中的表作为实验对象；在两个数据库中以相同的方式创建该表synctab，因为表结构不会自动复制

create table synctab(id int primary key,name text);

3.3 在所有节点设置允许Slony-I用户远程登录；在pg_hba.conf文件添加

host    all             slony             192.168.1.0/24        trust

3.4 设置slony(master主机操作)
编写一个slonik脚本用于注册这些节点的脚本如下所示：

[postgres@Postgres201 ~]$ cat slony_setup.sh 
#!/bin/sh
MASTERDB=lottu
SLAVEDB=lottu
HOST1=192.168.1.201
HOST2=192.168.1.202
DBUSER=slony
slonik<<_EOF_
cluster name = first_cluster;
# define nodes (this is needed by pretty much
# all slonik scripts)
node 1 admin conninfo = 'dbname=$MASTERDB host=$HOST1 user=$DBUSER';
node 2 admin conninfo = 'dbname=$SLAVEDB host=$HOST2 user=$DBUSER';
# init cluster
init cluster ( id=1, comment = 'Master Node');
# group tables into sets
create set (id=1, origin=1, comment='Our tables');
set add table (set id=1, origin=1, id=1, fully qualified name = 'lottu.synctab', comment='sample table');
store node (id=2, comment = 'Slave node', event node=1);
store path (server = 1, client = 2, conninfo='dbname=$MASTERDB host=$HOST1 user=$DBUSER');
store path (server = 2, client = 1, conninfo='dbname=$SLAVEDB host=$HOST2 user=$DBUSER');
_EOF_

现在这个表在Slony的控制下，我们可以开始订阅脚本如下所示：

[postgres@Postgres201 ~]$ cat slony_subscribe.sh 
#!/bin/sh
MASTERDB=lottu
SLAVEDB=lottu
HOST1=192.168.1.201
HOST2=192.168.1.202
DBUSER=slony
slonik<<_EOF_
cluster name = first_cluster;
node 1 admin conninfo = 'dbname=$MASTERDB host=$HOST1 user=$DBUSER';
node 2 admin conninfo = 'dbname=$SLAVEDB host=$HOST2 user=$DBUSER';
subscribe set ( id = 1, provider = 1, receiver = 2, forward = no);
_EOF_

在master主机执行脚本

[postgres@Postgres201 ~]$ ./slony_setup.sh 
[postgres@Postgres201 ~]$ ./slony_subscribe.sh &
[1] 1225

定义了我们想要复制的东西之后，我们可以在每台主机启动slon守护进程

slon first_cluster 'host=192.168.1.201 dbname=lottu user=slony' &
slon first_cluster 'host=192.168.1.202 dbname=lottu user=slony' &

验证slony-I是否配置成功。
在master主机执行dml操作

[postgres@Postgres201 ~]$ psql lottu lottu
psql (9.6.0)
Type "help" for help.

lottu=# \d synctab
    Table "lottu.synctab"
 Column |  Type   | Modifiers 
--------+---------+-----------
 id     | integer | not null
 name   | text    | 
Indexes:
    "synctab_pkey" PRIMARY KEY, btree (id)
Triggers:
    _first_cluster_logtrigger AFTER INSERT OR DELETE OR UPDATE ON synctab FOR EACH ROW EXECUTE PROCEDURE _first_cluster.logtrigger('_first_cluster', '1', 'k')
    _first_cluster_truncatetrigger BEFORE TRUNCATE ON synctab FOR EACH STATEMENT EXECUTE PROCEDURE _first_cluster.log_truncate('1')
Disabled user triggers:
    _first_cluster_denyaccess BEFORE INSERT OR DELETE OR UPDATE ON synctab FOR EACH ROW EXECUTE PROCEDURE _first_cluster.denyaccess('_first_cluster')
    _first_cluster_truncatedeny BEFORE TRUNCATE ON synctab FOR EACH STATEMENT EXECUTE PROCEDURE _first_cluster.deny_truncate()
lottu=# insert into synctab values (1001,'lottu');
INSERT 0 1

在slave主机查看是否对应变化

[postgres@Postgres202 ~]$ psql
psql (9.6.0)
Type "help" for help.

postgres=# \c lottu lottu
You are now connected to database "lottu" as user "lottu".
lottu=> select * from synctab ;
  id  | name  
------+-------
 1001 | lottu
(1 row)

4. Slony-I相关表或者视图查看

4.1 配置成功；会在所在的数据库中生成一个schema

[postgres@Postgres201 ~]$ psql lottu lottu
psql (9.6.0)
Type "help" for help.

lottu=# \dn
      List of schemas
      Name      |  Owner   
----------------+----------
 _first_cluster | slony
 lottu          | lottu
 public         | postgres
(3 rows)

4.2 查看集群中的节点信息

lottu=# select * from _first_cluster.sl_node;
 no_id | no_active | no_comment  | no_failed 
-------+-----------+-------------+-----------
     1 | t         | Master Node | f
     2 | t         | Slave node  | f
(2 rows)

4.3 查看集群中的集合信息

lottu=# select * from _first_cluster.sl_set;
 set_id | set_origin | set_locked | set_comment 
--------+------------+------------+-------------
      1 |          1 |            | Our tables
(1 row)

4.4 查看集群中的表信息

lottu=# select * from _first_cluster.sl_table;
 tab_id | tab_reloid | tab_relname | tab_nspname | tab_set | tab_idxname  | tab_altered | tab_comment  
--------+------------+-------------+-------------+---------+--------------+-------------+--------------
      1 |      57420 | synctab     | lottu       |       1 | synctab_pkey | f           | sample table
(1 row)

5. 日常维护

5.1 Slony-I向现有集群中增加一个复制表
以表synctab2为例：

create table synctab2(id int primary key,name text,reg_time timestamp);

我们要创建一个新的表格集;脚本是这样的

[postgres@Postgres201 ~]$ cat slony_add_table_set.sh 
#!/bin/sh
MASTERDB=lottu
SLAVEDB=lottu
HOST1=192.168.1.201
HOST2=192.168.1.202
DBUSER=slony
slonik<<_EOF_
cluster name = first_cluster;
node 1 admin conninfo = 'dbname=$MASTERDB host=$HOST1 user=$DBUSER';
node 2 admin conninfo = 'dbname=$SLAVEDB host=$HOST2 user=$DBUSER';
create set (id=2, origin=1, comment='a second replication set');
set add table (set id=2, origin=1, id=2, fully qualified name ='lottu.synctab2', comment='second table');
subscribe set(id=1, provider=1,receiver=2);
subscribe set(id=2, provider=1,receiver=2);
merge set(id=1, add id=2,origin=1);
_EOF_

执行slony_add_table_set.sh脚本

[postgres@Postgres201 ~]$ ./slony_add_table_set.sh 
<stdin>:8 subscription in progress before mergeSet. waiting
<stdin>:8 subscription in progress before mergeSet. waiting

查看是否添加成功

lottu=# select * from _first_cluster.sl_table;
 tab_id | tab_reloid | tab_relname | tab_nspname | tab_set |  tab_idxname  | tab_altered | tab_comment  
--------+------------+-------------+-------------+---------+---------------+-------------+--------------
      1 |      57420 | synctab     | lottu       |       1 | synctab_pkey  | f           | sample table
      2 |      57840 | synctab2    | lottu       |       1 | synctab2_pkey | f           | second table
(2 rows)

5.2 Slony-I向现有集群中删除一个复制表

[postgres@Postgres201 ~]$ cat slony_drop_table.sh
#!/bin/sh
MASTERDB=lottu
SLAVEDB=lottu
HOST1=192.168.1.201
HOST2=192.168.1.202
DBUSER=slony
slonik<<_EOF_
cluster name = first_cluster;
node 1 admin conninfo = 'dbname=$MASTERDB host=$HOST1 user=$DBUSER';
node 2 admin conninfo = 'dbname=$SLAVEDB host=$HOST2 user=$DBUSER';
set drop table (id=2, origin=1);
_EOF_

执行slony_drop_table.sh脚本

[postgres@Postgres201 ~]$ ./slony_drop_table.sh

查看是否删除成功

lottu=# select * from _first_cluster.sl_table;
 tab_id | tab_reloid | tab_relname | tab_nspname | tab_set | tab_idxname  | tab_altered | tab_comment  
--------+------------+-------------+-------------+---------+--------------+-------------+--------------
      1 |      57420 | synctab     | lottu       |       1 | synctab_pkey | f           | sample table
(1 row)

3 删除slony

[postgres@Postgres201 ~]$ cat slony_drop_node.sh 
#!/bin/sh
MASTERDB=lottu
SLAVEDB=lottu
HOST1=192.168.1.201
HOST2=192.168.1.202
DBUSER=slony
slonik<<_EOF_
cluster name = first_cluster;
node 1 admin conninfo = 'dbname=$MASTERDB host=$HOST1 user=$DBUSER';
node 2 admin conninfo = 'dbname=$SLAVEDB host=$HOST2 user=$DBUSER';
uninstall node (id = 1);
uninstall node (id = 2);
_EOF_

执行脚本如下

postgres@Postgres201 ~]$ ./slony_drop_node.sh 
<stdin>:4: NOTICE:  Slony-I: Please drop schema "_first_cluster"
<stdin>:4: NOTICE:  drop cascades to 175 other objects
.........
drop cascades to function _first_cluster.unlockset(integer)
drop cascades to function _first_cluster.moveset(integer,integer)
drop cascades to function _first_cluster.moveset_int(integer,integer,integer,bigint)
and 75 other objects (see server log for list)

完美；一切归零！

查考文献
https://www.cnblogs.com/ilifeilong/p/7009322.html
https://www.cnblogs.com/gaojian/p/3196244.html

PostgreSQL逻辑复制之slony篇