pacemaker+corosync 搭建一主两从PG集群
2024-04-10 05:44:04
一、OS配置
1.1环境规划
环境：centos 7
数据库：pg11.5
ip地址：
node1 10.122.166.120
node2 10.122.166.127
node3 10.122.166.128
vip：
10.122.166.123
10.122.166.133
1.2系统配置
关闭selinux ‐‐‐‐> /etc/selinux/config
关闭防火墙
systemctl stop firewalld.service
systemctl disable firewalld.service
修改主机名
hostnamectl set‐hostname nodeX
编辑 /etc/hosts 文件
10.122.166.120 node1
10.122.166.127 node2
10.122.166.128 node3
二、数据库安装
2.1 所有节点安装PG软件（以node1为例）
1)安装依赖包
[root@node1 ~]# yum install gcc gcc-c++ flex bison readline-devel zlib-devel -y2)创建用户
[root@node1 ~]# groupadd postgres
[root@node1 ~]# useradd postgres -g postgres
[root@node1 ~]# echo "pg123"|passwd postgres --stdin3)编译安装PG软件
[root@node1 ~]# tar -xvf postgresql-11.5.tar.gz
[root@node1 ~]# cd postgresql-11.5/
[root@node1 ~]# ./configure --with-segsize=20 --prefix=/data/postgres/
[root@node1 ~]# cd contrib/
[root@node1 ~]# gmake install all4)创建数据目录、归档目录
[root@node1 ~]# mkdir -p /data/postgres/data
[root@node1 ~]# mkdir -p /data/postgres/archive_log
[root@node1 ~]# chown -R postgres:postgres /data/postgres
[root@node1 ~]# chmod 0700 /data/postgres/data2.2 node1 初始化数据库
1)初始化数据库
[postgres@node1 ~]$ cd /data/postgres/bin/
[postgres@node1 bin]$ initdb ‐D /data/postgres/data2)修改配置文件
[postgres@node1 data]$ vim postgresql.conflisten_addresses = '*'
port = '5432'
max_connections = '1000'
log_connections = 'yes'
shared_buffers = '4000MB'
effective_cache_size = '1000MB'
temp_buffers = '64MB'
work_mem = '40MB'
max_worker_processes = '32'
max_parallel_workers = '16'
max_parallel_workers_per_gather = '4'
wal_compression = 'on'
log_destination = 'csvlog'
logging_collector = 'on'
log_directory = 'log'
log_truncate_on_rotation = 'on'
log_rotation_size = '30MB'
log_lock_waits = 'on'
lock_timeout = '300000'
wal_level = replica
synchronous_commit = on
archive_mode = on
archive_command = 'cp %p /data/postgres/archive_log/%f'
max_wal_senders=5
wal_keep_segments = 32
hot_standby = on
wal_log_hints = 'on'
log_statement = 'ddl'[postgres@node1 data]$ vim pg_hba.confhost all all 10.122.166.0/24 trust
host replication all 10.122.166.0/24 trust3)添加环境变量
[postgres@node1 ~]$ vi ~/.bash_profileexport PATH=/data/postgres/bin/:$PATH
export PGDATA=/data/postgres/data
export LD_LIBRARY_PATH=/data/postgres/lib:${LD_LIBRARY_PATH}[postgres@node1 ~]$ source ~/.bash_profile4)启动数据库
[postgres@node1 ~]$ pg_ctl start2.3 node2/node3 创建备库
1)以 node2 为例
[postgres@node2 data]$ pg_basebackup -h 10.122.166.120 -p 5432 -U postgres -D /data/postgres/data2)配置recovery.conf 文件
[postgres@node2 data]$ vim /data/postgres/data/recovery.conf
standby_mode = on
primary_conninfo = 'host=10.122.166.120 port=5432 user=postgres password=postgres'
recovery_target_timeline = 'latest'3)启动备库
[postgres@node2 ~]$ pg_ctl start2.4 node1 查看集群状态
postgres=# select application_name,client_addr,sync_state from pg_stat_replication;application_name | client_addr | sync_state
------------------+----------------+------------node2 | 10.122.166.127 | asyncnode3 | 10.122.166.128 | async
(2 rows)2.5 所有节点停止 PG服务
先备库 后主库
pg_ctl -D stop -m f三、配置corosync
1)所有节点安装软件
yum install -y pacemaker corosync pcs2)配置 corosync.conf文件
[root@node1 ~]# cd /etc/corosync
[root@node1 corosync]# vim corosync.conf
totem {max_messages: 20vsftype: nonecrypto_cipher: nonesecauth: ontoken_retransmits_before_loss_const: 10interface {bindnetaddr: 10.122.166.0mcastaddr: 226.94.61.22ringnumber: 0mcastport: 5405ttl: 1}consensus: 6000clear_node_high_bit: yescrypto_hash: nonerrp_mode: nonejoin: 60cluster_name: clustertoken: 5000version: 2transport: udpip_version: ipv4heartbeat_failures_allowed : 3
}
logging {to_logfile: yestimestamp: onsyslog_facility: daemonlogger_subsys {debug: offsubsys: QUORUM}to_syslog: yesdebug: offlogfile: /var/log/corosync.log to_stderr: nofileline: off
}
quorum {expected_votes: 3two_node: 0provider: corosync_votequorum
}nodelist {node {ring0_addr: node1nodeid: 1}node {ring0_addr: node2nodeid: 2}node {ring0_addr: node3nodeid: 3}
}3)复制到另外两个节点
[root@node1 corosync]# scp corosync.conf 10.122.166.127:/etc/corosync
[root@node1 corosync]# scp corosync.conf 10.122.166.128:/etc/corosync4)所有节点启动corosync服务
service corosync start四、配置 pacemaker
1)所有节点启动 pacemaker服务
service pacemaker start2)任意节点使用root用户检查状态
[root@node3 cib]# crm_mon -Afr -1
Stack: corosync
Current DC: node2 (version 1.1.23-1.el7_9.1-9acf116022) - partition with quorum
Last updated: Thu Sep 29 13:59:07 2022
Last change: Thu Sep 29 13:58:59 2022 by hacluster via crmd on node23 nodes configured
0 resource instances configuredOnline: [ node1 node2 node3 ]No resourcesNode Attributes:
* Node node1:
* Node node2:
* Node node3:Migration Summary:
* Node node2:
* Node node1:
* Node node3:3)node1 编辑config.pcs文件
[root@node1 ~]# cd /var/lib/pacemaker/
[root@node1 pacemaker]# vim config.pcspcs cluster cib pgsql_cfgpcs -f pgsql_cfg property set no-quorum-policy="ignore"
pcs -f pgsql_cfg property set stonith-enabled="false"
pcs -f pgsql_cfg resource defaults resource-stickiness="INFINITY"
pcs -f pgsql_cfg resource defaults migration-threshold="1"pcs -f pgsql_cfg resource create vip-master IPaddr2 \ip="10.122.166.123" \nic="eth0" \cidr_netmask="24" \op start timeout="60s" interval="0s" on-fail="restart" \op monitor timeout="60s" interval="10s" on-fail="restart" \op stop timeout="60s" interval="0s" on-fail="block"pcs -f pgsql_cfg resource create vip-slave IPaddr2 \ip="10.122.166.133" \nic="eth0" \cidr_netmask="24" \meta migration-threshold="0" \op start timeout="60s" interval="0s" on-fail="stop" \op monitor timeout="60s" interval="10s" on-fail="restart" \op stop timeout="60s" interval="0s" on-fail="ignore"pcs -f pgsql_cfg resource create pgsql pgsql \pgctl="/data/postgres/bin/pg_ctl" \psql="/data/postgres/bin/psql" \pgdata="/data/postgres/data/" \rep_mode="sync" \node_list="node1 node2 node3" \restore_command="cp %p /data/postgres/archive_log/%f" \primary_conninfo_opt="keepalives_idle=60 keepalives_interval=5 keepalives_count=5" \master_ip="10.122.166.123" \restart_on_promote='true' \op start timeout="60s" interval="0s" on-fail="restart" \op monitor timeout="60s" interval="4s" on-fail="restart" \op monitor timeout="60s" interval="3s" on-fail="restart" role="Master" \op promote timeout="60s" interval="0s" on-fail="restart" \op demote timeout="60s" interval="0s" on-fail="stop" \op stop timeout="60s" interval="0s" on-fail="block" \op notify timeout="60s" interval="0s"pcs -f pgsql_cfg resource master pgsql-cluster pgsql \master-max=1 master-node-max=1 clone-max=3 clone-node-max=1 notify=truepcs -f pgsql_cfg resource group add master-group vip-master
pcs -f pgsql_cfg resource group add slave-group vip-slavepcs -f pgsql_cfg constraint colocation add master-group with master pgsql-cluster INFINITY
pcs -f pgsql_cfg constraint order promote pgsql-cluster then start master-group symmetrical=false score=INFINITY
pcs -f pgsql_cfg constraint order demote pgsql-cluster then stop master-group symmetrical=false score=0 pcs -f pgsql_cfg constraint colocation add slave-group with slave pgsql-cluster INFINITY
pcs -f pgsql_cfg constraint order promote pgsql-cluster then start slave-group symmetrical=false score=INFINITY
pcs -f pgsql_cfg constraint order demote pgsql-cluster then stop slave-group symmetrical=false score=0 pcs cluster cib-push pgsql_cfg4)所有节点增加 root用户的 PATH路径
[root@node1 ~]# vim .bash_profile
export PATH=/data/postgres/bin/:$PATH
[root@node1 ~]# source .bash_profile5)node1 加载配置
[root@node1 pacemaker]# sh config.pcs
Warning: Defaults do not apply to resources which override them with their own defined values
Warning: Defaults do not apply to resources which override them with their own defined values
Assumed agent name 'ocf:heartbeat:IPaddr2' (deduced from 'IPaddr2')
Assumed agent name 'ocf:heartbeat:IPaddr2' (deduced from 'IPaddr2')
Assumed agent name 'ocf:heartbeat:pgsql' (deduced from 'pgsql')
Adding pgsql-cluster master-group (score: INFINITY) (Options: first-action=promote then-action=start symmetrical=false)
Adding pgsql-cluster master-group (score: 0) (Options: first-action=demote then-action=stop symmetrical=false)
Adding pgsql-cluster slave-group (score: INFINITY) (Options: first-action=promote then-action=start symmetrical=false)
Adding pgsql-cluster slave-group (score: 0) (Options: first-action=demote then-action=stop symmetrical=false)
CIB updated6)等待一会，再次检查状态
[root@node1 ~]# crm_mon -Afr -1
Stack: corosync
Current DC: node2 (version 1.1.23-1.el7_9.1-9acf116022) - partition with quorum
Last updated: Thu Sep 29 15:14:27 2022
Last change: Thu Sep 29 14:15:00 2022 by root via crm_attribute on node13 nodes configured
5 resource instances configuredOnline: [ node1 node2 node3 ]Full list of resources:Master/Slave Set: pgsql-cluster [pgsql]Masters: [ node1 ]Slaves: [ node2 node3 ]Resource Group: master-groupvip-master (ocf::heartbeat:IPaddr2): Started node1Resource Group: slave-groupvip-slave (ocf::heartbeat:IPaddr2): Started node2Node Attributes:
* Node node1:+ master-pgsql : 1000+ pgsql-data-status : LATEST+ pgsql-master-baseline : 000000002D000098+ pgsql-status : PRI
* Node node2:+ master-pgsql : 100+ pgsql-data-status : STREAMING|SYNC+ pgsql-status : HS:sync
* Node node3:+ master-pgsql : -INFINITY+ pgsql-data-status : STREAMING|ASYNC+ pgsql-status : HS:asyncMigration Summary:
* Node node2:
* Node node1:
* Node node3:7)任意节点使用 pcs命令检查状态
[root@node1 ~]# pcs status
Cluster name: cluster
Stack: corosync
Current DC: node2 (version 1.1.23-1.el7_9.1-9acf116022) - partition with quorum
Last updated: Thu Sep 29 15:15:08 2022
Last change: Thu Sep 29 14:15:00 2022 by root via crm_attribute on node13 nodes configured
5 resource instances configuredOnline: [ node1 node2 node3 ]Full list of resources:Master/Slave Set: pgsql-cluster [pgsql]Masters: [ node1 ]Slaves: [ node2 node3 ]Resource Group: master-groupvip-master (ocf::heartbeat:IPaddr2): Started node1Resource Group: slave-groupvip-slave (ocf::heartbeat:IPaddr2): Started node2Daemon Status:corosync: active/disabledpacemaker: active/disabledpcsd: inactive/disabled8)在 node1(主库)查看状态
postgres=# select application_name,client_addr,sync_state from pg_stat_replication;application_name | client_addr | sync_state
------------------+----------------+------------node2 | 10.122.166.127 | syncnode3 | 10.122.166.128 | async
(2 rows)安装 crmsh
[root@ceph1 ~]# wget -P /etc/yum.repos.d/ http://download.opensuse.org/repositories/network:/ha-clustering:/Stable/CentOS_CentOS-7/network:ha-clustering:Stable.repo
[root@ceph1 ~]# yum install -y crmsh
[root@ceph1 ~]# crm configure property stonith-enabled=false使用 crm 命令
[root@node1 ~]# crm status
Stack: corosync
Current DC: node2 (version 1.1.23-1.el7_9.1-9acf116022) - partition with quorum
Last updated: Thu Sep 29 15:18:04 2022
Last change: Thu Sep 29 14:15:00 2022 by root via crm_attribute on node13 nodes configured
5 resource instances configuredOnline: [ node1 node2 node3 ]Full list of resources:Master/Slave Set: pgsql-cluster [pgsql]Masters: [ node1 ]Slaves: [ node2 node3 ]Resource Group: master-groupvip-master (ocf::heartbeat:IPaddr2): Started node1Resource Group: slave-groupvip-slave (ocf::heartbeat:IPaddr2): Started node2常用命令
停止集群服务： service pacemaker stop (会自动停止PG)
启动集群服务： service pacemaker start(会自动启动PG)
查看集群状态： crm_mon -Afr -1
查看集群状态且启动刷新: crm_mon -Afr
刷新集群状态： crm resource cleanup pgsql-cluster
刷新某一个节点的状态： crm resource cleanup pgsql nodename
清理历史告警: crm resource cleanuppacemaker+corosync 搭建一主两从PG集群相关推荐
- 利用pcs+pacemaker+corosync实现（HA）高可用集群
实验环境搭建 创建一台操作系统是rhel7.6的虚拟机node,配置好网络仓库,解析,网卡设置,关闭火墙和selinux后封装 克隆node虚拟机,虚拟机域名为node1,node2,node3,主机 ...
- Docker高级篇笔记（七）在docker里面进行redis三主三从的集群案例搭建步骤
目录 搭建三主三从集群 数据读写存储 主从容错的切换迁移 主从扩容 主从缩容 搭建三主三从集群 1 先启动docker 2 拉取redis镜像 3 创建6个redis容器 docker run -d ...
- 汇总-13台虚拟机搭建一个高可用负载均衡集群架构
要求 用13台虚拟机搭建一个高可用负载均衡集群架构出来,并运行三个站点,具体需求如下. 设计你认为合理的架构,用visio把架构图画出来 搭建lnmp.tomcat+jdk环境 三个站点分别为:dis ...
- 云服务器搭建高可用keepalived+nginx+emqx集群
云服务器搭建高可用keepalived+nginx+emqx集群 一.高可用emqx集群搭建 1 单机搭建emqx 2 配置认证和鉴权插件 3 搭建emqx集群 二.nginx搭建负载均衡 1 ngi ...
- 从零开始搭建高可用RabbitMQ镜像模式集群
文章目录 RabbitMQ集群模式搭建 准备工作 选取任意一个节点作为master节点, 进行文件同步, 我这里选择138作为master节点 组成集群 配置镜像队列(设置镜像队列策略) 集群配置参数 ...
- 搭建hbase1.2.5完全分布式集群
简介 有一段时间,没写博客了,因为公司开发分布式调用链追踪系统,用到hbase,在这里记录一下搭建过程 1.集群如下: ip 主机名 角色 192.168.6.130 node1.jacky.com ...
- 在Windows Server 2012 R2中搭建SQL Server 2012故障转移集群
需要说明的是我们搭建的SQL Server故障转移集群(SQL Server Failover Cluster)是可用性集群,而不是负载均衡集群,其目的是为了保证服务的连续性和可用性,而不是为了提高服 ...
- 3主3从redis集群扩缩容配置案例说明
3主3从redis集群扩缩容配置案例说明 文章目录 3主3从redis集群扩缩容配置案例说明 关闭防火墙+启动docker后台服务 新建6个docker实例 进入容器redis-node-1并为6台机 ...
- 一周碎碎念，2021.11.7，两个MGR集群间还可以构建传统的主从复制通道吗
叨叨最近遇到的一些事以及见闻.思考. 1. GreatSQL编译环境Dockerfile更新了 用于构建GreatSQL编译环境的Dockerfile发现几个小瑕疵,于是更新了下. 利用Docker环 ...
最新文章
- 苹果新功能惹众怒，4000 多家组织和个人签署公开信，敦促苹果放弃“儿童安全”功能...
- maven 下载包冲突问题
- 涨跌因子计算器下载哪里下载_网易超级计算器appv1-网易超级计算器v1下载
- 计算机操作系统：存储器的管理
- ccf报数游戏java,CCF 201712-2 游戏 《超越自我，攀登顶峰》
- 【VLC-Android】Mac下编译vlc-android
- SAP Data Intelligence API执行出错的排错之道
- [Swift]LeetCode482. 密钥格式化 | License Key Formatting
- 作用域经典练习题(主要是这个图太大不能一起分享)
- Prometheus和Grafana的监控Linux服务器和MYSQL数据库
- 定了！苹果 10 亿美元收购英特尔调制解调器部门！
- 二维数组七行七列C语言,C语言中级教程 再谈数组-7.ppt
- Java Server Faces (JSF)页面转换与验证（二）
- 用两个栈实现一个队列用两个队列实现一个栈
- Web前端课程设计-疫情解读网页
- oracle学习笔记-转
- MarkDown 分割线
- PMP-5.组织结构类型
- 神奇英语语法系列（二）——名词性从句
- 基于Gitee搭建免费图床