PostgreSQL 9.6 keepalived主从部署

## 环境：

PostgreSQL版：9.6

角色                     OS                    IP
master                 CentOS7　　　10.100.12.73

slave                    CentOS7          10.100.12.74

vIP                                                10.100.12.63

 ## 主从安装postgresql

postgresql官网安装文档：https://www.postgresql.org/download/linux/redhat/

* Install the repository RPM:

yum -y install https://download.postgresql.org/pub/repos/yum/9.6/redhat/rhel-7-x86_64/pgdg-redhat96-9.6-3.noarch.rpm

* Install the client packages:

yum -y install postgresql96

* Optionally install the server packages:

yum -y install postgresql96-server postgresql96-devel

* Optionally initialize the database and enable automatic start:

/usr/pgsql-9.6/bin/postgresql96-setup initdb
mv /usr/lib/systemd/system/postgresql-9.6.service /usr/lib/systemd/system/postgresql.service
systemctl enable postgresql
暂时先不启动服务

把/usr/pgsql-9.6/bin 加入系统环境变量

tail /etc/profile
## PATH
export PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/pgsql-9.6/bin

重载环境变量

. /etc/profile

## master服务器配置

启动postgresql服务

systemctl start postgresql

创建同步用户repluser

su - postgres
psql
create role repluser login replication encrypted password 'yHJ7TXda9q9zzIsv';
postgres=# du    # 查看用户
q

编辑 /var/lib/pgsql/9.6/data/pg_hba.conf  新增下面两行

host    replication     repluser        10.100.12.74/32         md5
host    all             all             0.0.0.0/0           md5

mkdir -p /data/pgsql/archivedir; chown -R postgres:postgres /data/pgsql/archivedir

编辑 /var/lib/pgsql/9.6/data/postgresql.conf

listen_addresses = '*'		# what IP address(es) to listen on;

port = 5432				# (change requires restart)
max_connections = 512			# (change requires restart) #从库的 max_connections要大于主库

shared_buffers = 128MB			# min 128kB

dynamic_shared_memory_type = posix	# the default is the first option

wal_level = hot_standby			# minimal, replica, or logical  #热备模式

archive_mode = on               # enables archiving; off, on, or always #允许归档

archive_command = 'test ! -f /data/pgsql/archivedir/%f && cp %p /data/pgsql/archivedir/%f'		# command to use to archive a logfile segment

max_wal_senders = 8		# max number of walsender processes #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 = 1024		# in logfile segments, 16MB each; 0 disables

log_destination = 'stderr'		# Valid values are combinations of

logging_collector = on			# Enable capturing of stderr and csvlog

log_directory = 'pg_log'		# directory where log files are written,

log_filename = 'postgresql-%a.log'	# log file name pattern,

log_truncate_on_rotation = on		# If on, an existing log file with the

log_rotation_age = 1d			# Automatic rotation of logfiles will

log_rotation_size = 0			# Automatic rotation of logfiles will

log_line_prefix = '< %m > '			# special values:

log_timezone = 'PRC'

datestyle = 'iso, mdy'
timezone = 'PRC'

lc_messages = 'en_US.UTF-8'			# locale for system error message

lc_monetary = 'en_US.UTF-8'			# locale for monetary formatting
lc_numeric = 'en_US.UTF-8'			# locale for number formatting
lc_time = 'en_US.UTF-8'				# locale for time formatting

default_text_search_config = 'pg_catalog.english'

重启postgresql服务

systemctl restart postgresql

## slave服务器配置

mkdir -p /data/pgsql/archivedir; chown -R postgres:postgres /data/pgsql/archivedir

su - postgres
rm -rf /var/lib/pgsql/9.6/data/*  #开始没有启动从库服务，这一步可以省略
pg_basebackup -h 10.100.12.73 -U repluser -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 /var/lib/pgsql/9.6/data/recovery.conf

grep -v "^#" /var/lib/pgsql/9.6/data/recovery.conf

recovery_target_timeline = 'latest'
standby_mode = on
primary_conninfo = 'host=10.100.12.73 port=5432 user=repluser password=yHJ7TXda9q9zzIsv'		# e.g. 'host=localhost port=5432'
trigger_file = '/var/lib/pgsql/9.6/data/trigger.kenyon'     #主从切换时后的触发文件，即 touch /var/lib/pgsql/9.6/data/trigger.kenyon 就可切换主从，也可以使用命令 /usr/pgsql-9.6/bin/pg_ctl promote

配置postgresql.conf文件

listen_addresses = '*'		# what IP address(es) to listen on;

port = 5432				# (change requires restart)
max_connections = 1024			# (change requires restart) 一般从的最大链接要大于主的

shared_buffers = 128MB			# min 128kB

dynamic_shared_memory_type = posix	# the default is the first option

wal_level = hot_standby			# minimal, replica, or logical

archive_mode = on               # enables archiving; off, on, or always

archive_command = 'test ! -f /data/pgsql/archivedir/%f && cp %p /data/pgsql/archivedir/%f'		# command to use to archive a logfile segment

max_wal_senders = 8		# max number of walsender processes

wal_keep_segments = 1024		# in logfile segments, 16MB each; 0 disables

hot_standby = on			# "on" allows queries during recovery  #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	# max delay before canceling queries

wal_receiver_status_interval = 10s	# send replies at least this often #多久向主报告一次从的状态

hot_standby_feedback = on		# send info from standby to prevent #如果有错误的数据复制，是否向主进行反馈

log_destination = 'stderr'		# Valid values are combinations of

logging_collector = on			# Enable capturing of stderr and csvlog

log_directory = 'pg_log'		# directory where log files are written,

log_filename = 'postgresql-%a.log'	# log file name pattern,

log_truncate_on_rotation = on		# If on, an existing log file with the

log_rotation_age = 1d			# Automatic rotation of logfiles will

log_rotation_size = 0			# Automatic rotation of logfiles will

log_line_prefix = '< %m > '			# special values:

log_timezone = 'PRC'

datestyle = 'iso, mdy'
timezone = 'PRC'

lc_messages = 'en_US.UTF-8'			# locale for system error message

lc_monetary = 'en_US.UTF-8'			# locale for monetary formatting
lc_numeric = 'en_US.UTF-8'			# locale for number formatting
lc_time = 'en_US.UTF-8'				# locale for time formatting

default_text_search_config = 'pg_catalog.english'

启动 postgresql服务

## 查看postgresql主从状态

在 master上执行
su - postgres

psql

select client_addr,sync_state from pg_stat_replication;

select * from pg_stat_replication;

pg_controldata /var/lib/pgsql/9.6/data    # 这种方法对于直接kill进程的情况下是不适用的，查看结果不准确，Database cluster state：信息

主库状态为：in production

备机状态为： in archive recovery

## keepalived配置

主从安装keepalived

yum -y install keepalived

master keepalived 配置

cat /etc/keepalived/keepalived.conf

global_defs {
        notification_email {
                admin@xx.com
        }

        notification_email_from keepalived@xx.com
        smtp_server 127.0.0.1
        smtp_connect_timeout 30
        router_id pg_ha
}


vrrp_script chk_postgresql {
        script "/etc/keepalived/script/script/check_postgresql.sh |grep 'postgresql_success' "
        interval 2
        weight -10
}


vrrp_instance VI_1 {
        state BACKUP        ############ 辅机为 BACKUP
        interface eth0
        virtual_router_id 62
        mcast_src_ip 10.100.12.73
        priority 100                  ########### 权值要比 back 高
        advert_int 2
        nopreempt

        authentication {
                auth_type PASS
                auth_pass SNKQusp4kFpUKz
        }

        track_script { 
                chk_postgresql ### 执行监控的服务 
        }

        virtual_ipaddress {
                10.100.12.63
        }

        notify_master "/bin/python /etc/keepalived/script/keepalived_notify.py 'PostgreSQL-1 [10.100.12.73] change to master, vip:10.100.12.63' "
        notify_backup "/bin/python /etc/keepalived/script/keepalived_notify.py 'PostgreSQL-1 [10.100.12.73] postgresql check faild, change to slave, vip:10.100.12.63' "

}

sh脚本：

cd /etc/keepalived/script    #sh脚本赋予可执行权限

cat check_postgresql.sh

#!/bin/bash
# songyanlin

pguser="postgres"
BIN="/usr/pgsql-9.6/bin"
datef=`date +%Y-%M-%d" "%H:%m`
data_dir="/var/lib/pgsql/9.6/data"
log_dir="/var/log/postgresql.log"
service_name="postgresql"
pid="postmaster"
status="postgresql_failed"
status_success="postgresql_success"


function CheckService(){
    local ret=`$BIN/pg_controldata $data_dir |grep -E "in production|in archive recovery" |wc -l`
    echo $ret
}

function CheckPs(){
    local ret=`pidof $pid |wc -l`
    echo $ret
}

if [ $(CheckService) == 0 -o $(CheckPs) == 0 ]; then
    echo "$datef postgresql master status is erro!" >> $log_dir
    service $service_name restart
    if [ $(CheckService) != 0 -a $(CheckPs) != 0 ]; then
        status=$status_success
    fi
else
    status=$status_success
fi

echo $status

cat keepalived_notify.py

#!/usr/bin/env python
# -*- coding:utf-8 -*-

import smtplib
from email.mime.text import MIMEText
from email.header import Header
import sys, time, subprocess, random



# 第三方 SMTP 服务
mail_host="smtp.exmail.qq.com"  #设置服务器
userinfo_list = [{'user':'rp1@qq.com','pass':'pwd'}, {'user':'rp2@qq.com','pass':'pwd'}, {'user':'rp3@tuandai.com','pass':'pwd'}]

user_inst = userinfo_list[random.randint(0, len(userinfo_list)-1)]
mail_user=user_inst['user'] #用户名
mail_pass=user_inst['pass'] #口令


sender = mail_user # 邮件发送者
receivers = ['mymail@163.com', 'gogo@qq.com']  # 接收邮件，可设置为你的QQ邮箱或者其他邮箱


p = subprocess.Popen('hostname', shell=True, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
hostname = p.stdout.readline().split('
')[0]

message_to = ''
for i in receivers:
    message_to += i + ';'

def print_help():
    note = '''python script.py message
    '''
    print(note)
    exit(1)

time_stamp = time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time()))

message_content = ''
if len(sys.argv) == 2:
    message_content = '%s [%s] %s' %(time_stamp, hostname ,sys.argv[1])
    subject = '%s [%s] postgresql status is error' %(time_stamp, hostname)
else:
    print_help()

message = MIMEText(message_content, 'plain', 'utf-8')
message['From'] = Header(sender, 'utf-8')
message['To'] =  Header(message_to, 'utf-8')

message['Subject'] = Header(subject, 'utf-8')

try:
    smtpObj = smtplib.SMTP()
    smtpObj.connect(mail_host, 25)    # 25 为 SMTP 端口号
    smtpObj.login(mail_user,mail_pass)
    smtpObj.sendmail(sender, receivers, message.as_string())
    print("邮件发送成功")
except smtplib.SMTPException as e:
    print("Error: 无法发送邮件")
    print(e)

slave keepalived配置

cat /etc/keepalived/keepalived.conf

global_defs {
        notification_email {
                admin@xx.com
        }

        notification_email_from keepalived@xx.com
        smtp_server 127.0.0.1
        smtp_connect_timeout 30
        router_id pg_ha
}


vrrp_script chk_postgresql {
        script "/etc/keepalived/script/check_postgresql.sh |grep 'postgresql_success' "
        interval 2
        weight -10
}


vrrp_instance VI_1 {
        state BACKUP        ############ 辅机为 BACKUP
        interface eth0
        virtual_router_id 62
        mcast_src_ip 10.100.12.74
        priority 99                  ########### 权值要比 back 高
        advert_int 2
        #nopreempt

        authentication {
                auth_type PASS
                auth_pass SNKQusp4kFpUKz
        }

        track_script { 
                chk_postgresql ### 执行监控的服务 
        }

        virtual_ipaddress {
                10.100.12.63
        }

        notify_master "/etc/keepalived/script/postgresql_slave_to_master.sh"

}

sh脚本：

check_postgresql.sh  keepalived_notify.py与master相同

cat postgresql_slave_to_master.sh

#!/bin/bash
#

pguser="postgres"
BIN="/usr/pgsql-9.6/bin"
datef=`date +%Y-%M-%d" "%H:%m`
data_dir="/var/lib/pgsql/9.6/data"
log_dir="/var/log/postgresql.log"
service_name="postgresql"
pid="postmaster"
status="postgresql_failed"
status_success="postgresql_success"


function CheckService(){
    local ret=`$BIN/pg_controldata $data_dir |grep "in production" |wc -l`
    echo $tet
}

function CheckStatus(){
    local ret=`$BIN/pg_controldata $data_dir |grep "shut down in recovery" |wc -l`
    echo $ret
}

function CheckStatus2(){
    local ret=`$BIN/pg_controldata $data_dir |grep "in archive recovery" |wc -l`
    echo $ret
}


if [ $(CheckStatus) != 0 ];then
    service $service_name restart
fi

if [ $(CheckStatus2) != 0 ]; then
    su - $pguser -c "$BIN/pg_ctl promote"
fi

/bin/python /etc/keepalived/script/keepalived_notify.py "PostgreSQL[10.100.12.74] change to master, vip:10.100.12.63"

附：

若主从已经切换后，把原来的master设置为从，可按上面从机设置方法设置

postgresql扩展组件

报错：

Running handlers:
There was an error running gitlab-ctl reconfigure:

bash[migrate gitlab-rails database] (gitlab::database_migrations line 51) had an error: Mixlib::ShellOut::ShellCommandFailed: Expected process to exit with [0], but received '1'
---- Begin output of "bash"  "/tmp/chef-script20180125-31534-ul2ug1" ----
STDOUT: rake aborted!
ActiveRecord::StatementInvalid: PG::UndefinedFile: ERROR:  could not open extension control file "/usr/pgsql-9.6/share/extension/pg_trgm.control": No such file or directory
: CREATE EXTENSION IF NOT EXISTS "pg_trgm"
/opt/gitlab/embedded/service/gitlab-rails/db/schema.rb:18:in `block in <top (required)>'
/opt/gitlab/embedded/service/gitlab-rails/db/schema.rb:14:in `<top (required)>'
/opt/gitlab/embedded/service/gitlab-rails/lib/tasks/gitlab/db.rake:52:in `block (3 levels) in <top (required)>'
/opt/gitlab/embedded/bin/bundle:23:in `load'
/opt/gitlab/embedded/bin/bundle:23:in `<main>'

Caused by:
PG::UndefinedFile: ERROR:  could not open extension control file "/usr/pgsql-9.6/share/extension/pg_trgm.control": No such file or directory
/opt/gitlab/embedded/service/gitlab-rails/db/schema.rb:18:in `block in <top (required)>'
/opt/gitlab/embedded/service/gitlab-rails/db/schema.rb:14:in `<top (required)>'
/opt/gitlab/embedded/service/gitlab-rails/lib/tasks/gitlab/db.rake:52:in `block (3 levels) in <top (required)>'
/opt/gitlab/embedded/bin/bundle:23:in `load'
/opt/gitlab/embedded/bin/bundle:23:in `<main>'
Tasks: TOP => db:schema:load
(See full trace by running task with --trace)
-- enable_extension("plpgsql")
   -> 0.0224s
-- enable_extension("pg_trgm")
STDERR: 
---- End output of "bash"  "/tmp/chef-script20180125-31534-ul2ug1" ----
Ran "bash"  "/tmp/chef-script20180125-31534-ul2ug1" returned 1

yum -y install postgresql96-contrib-9.6.6    # 默认的 yum -y install postgresql-contrib

su - postgres

-bash-4.2$ psql gitlabhq_production
psql (9.6.6)
Type "help" for help.

postgres=# CREATE EXTENSION pg_trgm;