PostgreSQL DBA(31) – Backup&Recovery#4(搭建流复制)

PostgreSQL通过流复制Streaming Replication可轻松实现高可用HA环境的搭建.本节简单介绍了搭建流复制环境的基本步骤.

Step 1 主库:创建用户
创建复制用户replicator

testdb=# CREATE USER replicator WITH REPLICATION ENCRYPTED PASSWORD 'replicator';CREATE ROLE

Step 2 主库:参数配置
配置archive_mode等参数

archive_mode = ONwal_level = replicamax_wal_senders = 10archive_command = '/home/xdb/archive.sh %p %f'listen_addresses = '*'

也可用alter system命令修改

ALTER SYSTEM SET wal_level TO 'replica';ALTER SYSTEM SET archive_mode TO 'ON';ALTER SYSTEM SET max_wal_senders TO '10';ALTER SYSTEM SET listen_addresses TO '*';

重启数据库

pg_ctl -D $PGDATA restart -mf

Step 3 主库:访问配置
修改pg_hba.conf文件

host replication replicator 192.168.26.26/32 md5

生效配置

pg_ctl -D $PGDATA reload

Step 4 从库:从主库备份中恢复
在从库上使用pg_basebackup创建备库
192.168.26.25是主库IP,192.168.26.26是从库IP

pg_basebackup -h 192.168.26.25 -U replicator -p 5432 -D $PGDATA -P -Xs -R

配置从库postgres.conf

hot_standby = ONhot_standby_feedback = ONALTER SYSTEM SET hot_standby TO 'ON';ALTER SYSTEM SET hot_standby_feedback TO 'ON';

配置从库recovery.conf

$ cat $PGDATA/recovery.confstandby_mode = 'on'primary_conninfo = 'host=192.168.26.25 port=5432 user=replicator password=replicator'restore_command = 'cp /data/archivelog/%f %p'archive_cleanup_command = 'pg_archivecleanup /data/archivelog %r'

Step 5 从库:启动数据库

[xdb@localhost testdb]$ pg_ctl -D $PGDATA startwaiting for server to start....2019-03-13 12:13:30.239 CST [1870] LOG:  listening on IPv4 address "0.0.0.0", port 54322019-03-13 12:13:30.239 CST [1870] LOG:  listening on IPv6 address "::", port 54322019-03-13 12:13:30.252 CST [1870] LOG:  listening on Unix socket "/tmp/.s.PGSQL.5432"2019-03-13 12:13:30.379 CST [1870] LOG:  redirecting log output to logging collector process2019-03-13 12:13:30.379 CST [1870] HINT:  Future log output will appear in directory "pg_log". doneserver started

Step 6 验证复制环境
确认相关进程是否已启动

#主库[xdb@localhost testdb]$ ps -ef|grep senderxdb       1646  1532  0 12:13 ?        00:00:00 postgres: walsender replicator 192.168.26.26(35294) streaming 0/43000140xdb       1659  1440  0 12:17 pts/1    00:00:00 grep --color=auto sender[xdb@localhost testdb]$ #从库[xdb@localhost testdb]$ ps -ef|grep receiverxdb       1879  1870  0 12:13 ?        00:00:00 postgres: walreceiver   streaming 0/43000140xdb       1884  1799  0 12:18 pts/0    00:00:00 grep --color=auto receiver[xdb@localhost testdb]$ ps -ef|grep startupxdb       1872  1870  0 12:13 ?        00:00:00 postgres: startup   recovering 000000100000000000000043xdb       1887  1799  0 12:18 pts/0    00:00:00 grep --color=auto startup[xdb@localhost testdb]$

Step 7 监控
查询pg_stat_replication数据字典表

testdb=# SELECT * FROM pg_stat_replication;-[ RECORD 1 ]----+------------------------------pid              | 1646usesysid         | 90113usename          | replicatorapplication_name | walreceiverclient_addr      | 192.168.26.26client_hostname  | client_port      | 35294backend_start    | 2019-03-13 12:13:30.852269+08backend_xmin     | state            | streamingsent_lsn         | 0/43000140write_lsn        | 0/43000140flush_lsn        | 0/43000140replay_lsn       | 0/43000140write_lag        | flush_lag        | replay_lag       | sync_priority    | 0sync_state       | asynctestdb=#

同步复制
从库配置参数recovery.conf,在primary_conninfo中添加application_name

primary_conninfo = 'user=replicator password=replicator host=192.168.26.25 port=5432 application_name = standby_26'

主库配置参数

synchronous_standby_names = 'standby_26'synchronous_commit = on

重启数据库,验证是否配置成功

testdb=# \xExpanded display is on.testdb=# SELECT * FROM pg_stat_replication;-[ RECORD 1 ]----+------------------------------pid              | 2257usesysid         | 90113usename          | replicatorapplication_name | standby_26client_addr      | 192.168.26.26client_hostname  | client_port      | 35418backend_start    | 2019-03-13 15:17:57.330573+08backend_xmin     | 634state            | streamingsent_lsn         | 0/54D4DBD0write_lsn        | 0/54D4DBD0flush_lsn        | 0/54D4DBD0replay_lsn       | 0/54D4DBD0write_lag        | 00:00:00.00101flush_lag        | 00:00:00.001954replay_lag       | 00:00:00.002145sync_priority    | 1sync_state       | sync