本節簡單介紹了PostgreSQL 12搭建流複製的過程。PG 12把recovery.conf廢掉了，與PG 11略有不同，但大同小異。

**主庫**
建立複製使用者
```
[pg12@localhost pg120db]$ psql -c "CREATE USER replicator WITH REPLICATION ENCRYPTED PASSWORD 'test'" -d testdb
Timing is on.
Expanded display is used automatically.
CREATE ROLE
Time: 30.796 ms

```
常規引數配置
```
[pg12@localhost pg120db]$ grep 'listen' postgresql.conf 
listen_addresses = '*'		# what IP address(es) to listen on;

pg12@localhost pg120db]$ grep 'replication' pg_hba.conf 
# DATABASE can be "all", "sameuser", "samerole", "replication", a
# keyword does not match "replication". Access to replication
# "all", "sameuser", "samerole" or "replication" makes the name lose
# Allow replication connections from localhost, by a user with the
# replication privilege.
local   replication     all                                     trust
host    replication     all             127.0.0.1/32            trust
host    replication     all             192.168.0.0/16          md5
host    replication     all             ::1/128                 trust
[pg12@localhost pg120db]$

```
重新載入配置引數
```
[pg12@localhost pg120db]$ psql -c "select pg_reload_conf()" -d testdb
Timing is on.
Expanded display is used automatically.
 pg_reload_conf 
----------------
 t
(1 row)

Time: 454.580 ms

```
確認該庫為master主庫
```
[pg12@localhost pg120db]$ psql -c "select pg_is_in_recovery()" -d testdb
Timing is on.
Expanded display is used automatically.
 pg_is_in_recovery 
-------------------
 f
(1 row)

Time: 23.530 ms
[pg12@localhost pg120db]$

```

**備庫**
使用pg_basebackup執行主庫備份
```
[pg12@localhost ~]$ pg_basebackup -h 192.168.26.28 -U replicator -p 5432 -D $PGDATA -Fp -Xs -P -R
Password: 
 426401/9113562 kB (4%), 0/1 tablespace

```
其中-Fp表示以plain格式資料，-Xs表示以stream方式包含所需的WAL檔案，-P表示顯示進度，-R表示為replication寫配置資訊。
備份完成，使用-R選項，在data目錄下自動生成standby.signal“訊號”檔案（可手工使用touch命令生成）以及更新了postgresql.auto.conf檔案，postgresql.auto.conf中寫入了主庫的連線資訊（可手工新增primary_conninfo資訊）。
```
[pg12@localhost ~]$ pg_basebackup -h 192.168.26.28 -U replicator -p 5432 -D $PGDATA -Fp -Xs -P -R
Password: 
9113571/9113571 kB (100%), 1/1 tablespace
[pg12@localhost ~]$ 
[pg12@localhost ~]$ cd $PGDATA
[pg12@localhost testdb]$ ls
backup_label      pg_commit_ts   pg_log        pg_replslot   pg_stat_tmp  PG_VERSION            postgresql.conf
base              pg_dynshmem    pg_logical    pg_serial     pg_subtrans  pg_wal                standby.signal
current_logfiles  pg_hba.conf    pg_multixact  pg_snapshots  pg_tblspc    pg_xact
global            pg_ident.conf  pg_notify     pg_stat       pg_twophase  postgresql.auto.conf
[pg12@localhost testdb]$ ll standby.signal 
-rw-------. 1 pg12 pg12 0 Nov 12 16:35 standby.signal
[pg12@localhost testdb]$

[pg12@localhost testdb]$ cat postgresql.auto.conf
# Do not edit this file manually!
# It will be overwritten by the ALTER SYSTEM command.
primary_conninfo = 'user=replicator password=test host=192.168.26.28 port=5432 sslmode=prefer sslcompression=0 gssencmode=prefer krbsrvname=postgres target_session_attrs=any'
[pg12@localhost testdb]$

[pg12@localhost testdb]$ grep 'primary_conninfo' postgresql.*
postgresql.auto.conf:primary_conninfo = 'user=replicator password=test host=192.168.26.28 port=5432 sslmode=prefer sslcompression=0 gssencmode=prefer krbsrvname=postgres target_session_attrs=any'
postgresql.conf:#primary_conninfo = ''			# connection string to sending server
[pg12@localhost testdb]$

```
在PG 11中，需要建立recovery.conf檔案，在此檔案中配置standby_mode和primary_conninfo引數，PG 12已不再需要該檔案，改為standby.signal檔案以及透過配置引數直接設定。

備庫啟動資料庫，透過pg_is_in_recovery確認是否正常配置
```
[pg12@localhost testdb]$ pg_ctl -D $PGDATA start
waiting for server to start....2019-11-12 16:46:31.635 CST [20436] LOG:  starting PostgreSQL 12.0 on x86_64-pc-linux-gnu, compiled by gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-36), 64-bit
2019-11-12 16:46:31.636 CST [20436] LOG:  listening on IPv4 address "0.0.0.0", port 5432
2019-11-12 16:46:31.636 CST [20436] LOG:  listening on IPv6 address "::", port 5432
2019-11-12 16:46:31.638 CST [20436] LOG:  listening on Unix socket "/tmp/.s.PGSQL.5432"
2019-11-12 16:46:31.750 CST [20436] LOG:  redirecting log output to logging collector process
2019-11-12 16:46:31.750 CST [20436] HINT:  Future log output will appear in directory "pg_log".
 done
server started
[pg12@localhost testdb]$ psql -c "select pg_is_in_recovery()" -d testdb
 pg_is_in_recovery 
-------------------
 t
(1 row)

[pg12@localhost testdb]$

```

Time: 149.682 ms
[pg12@localhost pg120db]$

```

下一節將介紹如何使用HAproxy配置負載均衡。

**參考資料**
[How to Set Up Streaming Replication in PostgreSQL 12](https://www.percona.com/blog/2019/10/11/how-to-set-up-streaming-replication-in-postgresql-12/)

PostgreSQL DBA(126) - PG 12（搭建流複製）

相關文章