搭建高可用mysql系列（2）-- Percona XtraDB Cluster 安装
本文主要介绍在 centos 下 Percona XtraDB Cluster（下文简称PXC） 的安装, 个人的系统版本信息如下：
[root@c2-d09 worker]# more /etc/redhat-release
CentOS Linux release 7.3.1611 (Core)
一 安装前的准备工作
（1） 关闭 SELinux
这是因为SELinux 可能会限制 PXC 的数据访问,关闭的方法为命令行下执行：
setenforce 0
为了防止系统重启后上述操作实效，还需要修改 /etc/selinux/config 文件，将 SELINUX 设置为：permissive 即：
SELINUX=permissive
（2）确保 PXC 默认使用的的4个端口没有被防火墙阻止：
3306 mysql 实例端口
4444：用于SST传送的端口 ，可以在配置中进行修改，比如：wsrep_sst_receive_address=10.11.12.205:5555
4567：pxc cluster相互通讯的端口，可以在配置中进行修改，比如：wsrep_provider_options ="gmcast.listen_addr=tcp://0.0.0.0:4010; "
4568：用于IST传送的端口，可以在配置中进行修改，比如：wsrep_provider_options = "ist.recv_addr=10.11.12.206:7777; "
二 安装的具体步骤
（1）下载安装文件并解压到安装目录，本文的目录为：/home/worker/XtraDB_Cluster/XtraDB_3306
wget https://www.percona.com/downloads/Percona-XtraDB-Cluster-57/Percona-XtraDB- ,!Cluster-5.7.14-26.17/binary/tarball/Percona-XtraDB-Cluster-5.7.14-rel8-26.17.1. ,!Linux.x86_64.ssl101.tar.gz
（2）初始化第一个节点
（a）切换到安装目录，生成日志所需文件夹
cd /home/worker/XtraDB_Cluster/XtraDB_3306
mkdir -p logs/bin_log logs/innodb_log logs/relay_log var
（b）编辑 my.cnf 文件，本文用到的第一个节点 my.cnf 文件如下：
[client]
port = 3306
socket = /home/worker/XtraDB_Cluster/XtraDB_3306/var/mysql.sock [mysqld]
bind_address = 0.0.0.0
port = 3306
socket = /home/worker/XtraDB_Cluster/XtraDB_3306/var/mysql.sock
user = worker
datadir = /home/worker/XtraDB_Cluster/XtraDB_3306/data
pid_file = /home/worker/XtraDB_Cluster/XtraDB_3306/var/mysqld.pid
skip_name_resolve gtid_mode = ON
enforce_gtid_consistency = ON tmpdir = /tmp
slave_load_tmpdir = /tmp back_log = 1024
max_connections = 4096
max_connect_errors = 4096
table_open_cache = 1024
max_allowed_packet = 1M tmp_table_size = 128M
max_heap_table_size = 64M
read_buffer_size = 16M
read_rnd_buffer_size = 16M
sort_buffer_size = 32M
join_buffer_size = 8M
query_cache_size = 128M
query_cache_limit = 32K
open_files_limit = 65535 table_open_cache = 1024
table_definition_cache = 4096 thread_cache_size = 256
thread_stack = 192K
transaction_isolation = REPEATABLE-READ ft_min_word_len = 4
# Set if mysql only or small db
#memlock # Common logs
log_error = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/error.log
log_warnings
slow_query_log = 1
slow_query_log_file = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/slow.log
long_query_time = 2 general_log = 0
general_log_file = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/general.log # Replication
server_id = 12616010
#log-slave-updates = 1
log_bin = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/bin_log/mysql-bin
log_bin_index = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/bin_log/mysql-bin.index
binlog_format = ROW
binlog_cache_size = 32M
expire_logs_days = 31
sync_binlog = 1
log_slave_updates = 1 relay_log = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/relay_log/mysql-relay-bin
relay_log_index = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/relay_log/mysql-relay-bin.index
relay_log_info_file = relay-log.info
master-info-file = master.info skip_slave_start = 1 #*** MyISAM Specific options
key_buffer_size = 128M
bulk_insert_buffer_size = 64M
myisam_sort_buffer_size = 32M
myisam_max_sort_file_size = 128M
myisam_repair_threads = 1
myisam-recover-options # *** INNODB Specific options ***
innodb_file_per_table = 1
innodb_buffer_pool_size = 4G
innodb_data_home_dir = /home/worker/XtraDB_Cluster/XtraDB_3306/data
innodb_data_file_path = ibdata1:256M:autoextend
innodb_max_dirty_pages_pct = 90 innodb_log_buffer_size = 16M
innodb_log_file_size = 256M
innodb_log_files_in_group = 3
innodb_log_group_home_dir = /home/worker/XtraDB_Cluster/XtraDB_3306/logs/innodb_log innodb_flush_method = O_DIRECT innodb_write_io_threads = 8
innodb_read_io_threads = 8 innodb_flush_log_at_trx_commit = 1 # Speed up InnoDB shutdown. This will disable InnoDB to do a full purge
# and insert buffer merge on shutdown. It may increase shutdown time a
# lot, but InnoDB will have to do it on the next startup instead.
#innodb_fast_shutdown innodb_lock_wait_timeout = 120 #pxc
default_storage_engine=Innodb
innodb_locks_unsafe_for_binlog=1
innodb_autoinc_lock_mode=2
pxc_strict_mode=ENFORCING wsrep_cluster_address=gcomm://10.126.160.1:33060,10.126.160.2:33070,10.126.160.3:33080
wsrep_provider_options="gmcast.listen_addr=tcp://0.0.0.0:33060;ist.recv_addr=10.126.160.1:33061"
wsrep_sst_receive_address=10.126.160.1:33062 wsrep_cluster_name=pxc_default_channel
wsrep_node_name=defaultchannel_126_160_1_3306
wsrep_provider=/home/worker/XtraDB_Cluster/XtraDB_3306/lib/libgalera_smm.so
wsrep_sst_method=xtrabackup-v2
wsrep_sst_auth="sst:defaultchannel"
wsrep_node_address=10.126.160.1
wsrep_slave_threads=8 [mysqldump]
quick max_allowed_packet = 1M [mysql]
no_auto_rehash # Only allow UPDATEs and DELETEs that use keys.
#safe_updates [myisamchk]
key_buffer_size = 512M
sort_buffer_size = 512M
read_buffer = 8M
write_buffer = 8M [mysqlhotcopy]
interactive_timeout [mysqld_safe]
open_files_limit = 65535
（c）初始化mysql
sudo ./bin/mysqld --defaults-file=/home/worker/XtraDB_Cluster/XtraDB_3306/my.cnf --basedir=/home/worker/XtraDB_Cluster/XtraDB_3306 --initialize-insecure --user=worker
（d）加载数据并启动mysql
如果有使用 mysql dump 等方式导出的数据，可以将数据copy 到 /home/worker/XtraDB_Cluster/XtraDB_3306/data 目录下
./bin/mysqld_safe --defaults-file=/home/worker/XtraDB_Cluster/XtraDB_3306/my.cnf
（e）启动PXC第一个节点
nohup ./bin/mysqld_safe --defaults-file=/home/worker/XtraDB_Cluster/XtraDB_3306/my.cnf --wsrep-new-cluster &
（f）验证节点是否正常启动
使用 socket 进入mysql： ./bin/mysql --socket=./var/mysql.sock -uroot 然后执行： show status like 'wsrep%'; 得到如下的结果，代表第一个节点已经正常启动：
+------------------------------+-------------------------------------------------------+
| Variable_name | Value |
+------------------------------+-------------------------------------------------------+
| wsrep_local_state_uuid | 3afab882-07ef-11e8-b57a-ce630de1af49 |
| ... | ...
| wsrep_local_state | 4 |
| wsrep_local_state_comment | Synced |
| ... |
| wsrep_cluster_size | 1 |
| wsrep_cluster_status | Primary |
| wsrep_connected | ON |
| ... | ...
| wsrep_ready | ON |
+------------------------------+-------------------------------------------------------+
59 rows in set (0.01 sec)
（g）创建sst用户，用于数据同步
在第一个节点 mysql下执行以下命令： CREATE USER 'sst' IDENTIFIED BY 'defaultchannel'; GRANT RELOAD, LOCK TABLES, PROCESS, REPLICATION CLIENT ON *.* TO 'sst'; FLUSH PRIVILEGES;
至此，第一个节点创建完毕。
（3） 初始化其它节点
(a) 切换到安装目录，生成日志所需文件夹
跟步骤（2）中 （a）步骤一致
（b）编辑 my.cnf 文件，本文用到的第二，第三节点的 my.cnf 文件如下：
第二节点：
[client]
port = 3307
socket = /home/worker/XtraDB_Cluster/XtraDB_3307/var/mysql.sock [mysqld]
bind_address = 0.0.0.0
port = 3307
socket = /home/worker/XtraDB_Cluster/XtraDB_3307/var/mysql.sock
user = worker
datadir = /home/worker/XtraDB_Cluster/XtraDB_3307/data
pid_file = /home/worker/XtraDB_Cluster/XtraDB_3307/var/mysqld.pid
skip_name_resolve gtid_mode = ON
enforce_gtid_consistency = ON tmpdir = /tmp
slave_load_tmpdir = /tmp back_log = 1024
max_connections = 4096
max_connect_errors = 4096
table_open_cache = 1024
max_allowed_packet = 1M tmp_table_size = 128M
max_heap_table_size = 64M
read_buffer_size = 16M
read_rnd_buffer_size = 16M
sort_buffer_size = 32M
join_buffer_size = 8M
query_cache_size = 128M
query_cache_limit = 32K
open_files_limit = 65535 table_open_cache = 1024
table_definition_cache = 4096 thread_cache_size = 256
thread_stack = 192K
transaction_isolation = REPEATABLE-READ ft_min_word_len = 4
# Set if mysql only or small db
#memlock # Common logs
log_error = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/error.log
log_warnings
slow_query_log = 1
slow_query_log_file = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/slow.log
long_query_time = 2 general_log = 0
general_log_file = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/general.log # Replication
server_id = 12616021
#log-slave-updates = 1 log_bin = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/bin_log/mysql-bin
log_bin_index = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/bin_log/mysql-bin.index
binlog_format = ROW
binlog_cache_size = 32M
expire_logs_days = 31
sync_binlog = 1
log_slave_updates = 1 relay_log = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/relay_log/mysql-relay-bin
relay_log_index = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/relay_log/mysql-relay-bin.index
relay_log_info_file = relay-log.info
master-info-file = master.info skip_slave_start = 1 #*** MyISAM Specific options
key_buffer_size = 128M
bulk_insert_buffer_size = 64M
myisam_sort_buffer_size = 32M
myisam_max_sort_file_size = 128M
myisam_repair_threads = 1
myisam-recover-options # *** INNODB Specific options ***
innodb_file_per_table = 1
innodb_buffer_pool_size = 4G
innodb_data_home_dir = /home/worker/XtraDB_Cluster/XtraDB_3307/data
innodb_data_file_path = ibdata1:256M:autoextend
innodb_max_dirty_pages_pct = 90 innodb_log_buffer_size = 16M
innodb_log_file_size = 256M
innodb_log_files_in_group = 3
innodb_log_group_home_dir = /home/worker/XtraDB_Cluster/XtraDB_3307/logs/innodb_log innodb_flush_method = O_DIRECT innodb_write_io_threads = 8
innodb_read_io_threads = 8 innodb_flush_log_at_trx_commit = 1 # Speed up InnoDB shutdown. This will disable InnoDB to do a full purge
# and insert buffer merge on shutdown. It may increase shutdown time a
# lot, but InnoDB will have to do it on the next startup instead.
#innodb_fast_shutdown innodb_lock_wait_timeout = 120 #pxc
default_storage_engine=Innodb
innodb_locks_unsafe_for_binlog=1
innodb_autoinc_lock_mode=2
pxc_strict_mode=ENFORCING wsrep_cluster_address=gcomm://10.126.160.1:33060,10.126.160.2:33070,10.126.160.3:33080
wsrep_provider_options="gmcast.listen_addr=tcp://0.0.0.0:33070;ist.recv_addr=10.126.160.2:33071"
wsrep_sst_receive_address=10.126.160.2:33072
wsrep_cluster_name=pxc_default_channel
wsrep_node_name=defaultchannel_126_160_2_3307
wsrep_provider=/home/worker/XtraDB_Cluster/XtraDB_3307/lib/libgalera_smm.so
wsrep_sst_method=xtrabackup-v2
wsrep_sst_auth="sst:defaultchannel"
wsrep_node_address=10.126.160.2
wsrep_slave_threads=8 [mysqldump]
quick max_allowed_packet = 1M [mysql]
no_auto_rehash # Only allow UPDATEs and DELETEs that use keys.
#safe_updates [myisamchk]
key_buffer_size = 512M
sort_buffer_size = 512M
read_buffer = 8M
write_buffer = 8M [mysqlhotcopy]
interactive_timeout [mysqld_safe]
open_files_limit = 65535
第三个节点：
[client]
port = 3308
socket = /home/worker/XtraDB_Cluster/XtraDB_3308/var/mysql.sock [mysqld]
bind_address = 0.0.0.0
port = 3308
socket = /home/worker/XtraDB_Cluster/XtraDB_3308/var/mysql.sock
user = worker
datadir = /home/worker/XtraDB_Cluster/XtraDB_3308/data
pid_file = /home/worker/XtraDB_Cluster/XtraDB_3308/var/mysqld.pid
skip_name_resolve gtid_mode = ON
enforce_gtid_consistency = ON tmpdir = /tmp
slave_load_tmpdir = /tmp back_log = 1024
max_connections = 4096
max_connect_errors = 4096
table_open_cache = 1024
max_allowed_packet = 1M tmp_table_size = 128M
max_heap_table_size = 64M
read_buffer_size = 16M
read_rnd_buffer_size = 16M
sort_buffer_size = 32M
join_buffer_size = 8M
query_cache_size = 128M
query_cache_limit = 32K
open_files_limit = 65535 table_open_cache = 1024
table_definition_cache = 4096 thread_cache_size = 256
thread_stack = 192K
transaction_isolation = REPEATABLE-READ ft_min_word_len = 4
# Set if mysql only or small db
#memlock # Common logs
log_error = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/error.log
log_warnings
slow_query_log = 1
slow_query_log_file = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/slow.log
long_query_time = 2 general_log = 0
general_log_file = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/general.log # Replication
server_id = 12616032
#log-slave-updates = 1 log_bin = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/bin_log/mysql-bin
log_bin_index = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/bin_log/mysql-bin.index
binlog_format = ROW
binlog_cache_size = 32M
expire_logs_days = 31
sync_binlog = 1
log_slave_updates = 1 relay_log = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/relay_log/mysql-relay-bin
relay_log_index = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/relay_log/mysql-relay-bin.index
relay_log_info_file = relay-log.info
master-info-file = master.info skip_slave_start = 1 #*** MyISAM Specific options
key_buffer_size = 128M
bulk_insert_buffer_size = 64M
myisam_sort_buffer_size = 32M
myisam_max_sort_file_size = 128M
myisam_repair_threads = 1
myisam-recover-options # *** INNODB Specific options ***
innodb_file_per_table = 1
innodb_buffer_pool_size = 4G
innodb_data_home_dir = /home/worker/XtraDB_Cluster/XtraDB_3308/data
innodb_data_file_path = ibdata1:256M:autoextend
innodb_max_dirty_pages_pct = 90 innodb_log_buffer_size = 16M
innodb_log_file_size = 256M
innodb_log_files_in_group = 3
innodb_log_group_home_dir = /home/worker/XtraDB_Cluster/XtraDB_3308/logs/innodb_log innodb_flush_method = O_DIRECT innodb_write_io_threads = 8
innodb_read_io_threads = 8 innodb_flush_log_at_trx_commit = 1 # Speed up InnoDB shutdown. This will disable InnoDB to do a full purge
# and insert buffer merge on shutdown. It may increase shutdown time a
# lot, but InnoDB will have to do it on the next startup instead.
#innodb_fast_shutdown innodb_lock_wait_timeout = 120 #pxc
default_storage_engine=Innodb
innodb_locks_unsafe_for_binlog=1
innodb_autoinc_lock_mode=2
pxc_strict_mode=ENFORCING wsrep_cluster_address=gcomm://10.126.160.1:33060,10.126.160.2:33070,10.126.160.3:33080
wsrep_provider_options="gmcast.listen_addr=tcp://0.0.0.0:33080;ist.recv_addr=10.126.160.3:33081"
wsrep_sst_receive_address=10.126.160.3:33082 wsrep_cluster_name=pxc_default_channel
wsrep_node_name=defaultchannel_126_160_3_3308
wsrep_provider=/home/worker/XtraDB_Cluster/XtraDB_3308/lib/libgalera_smm.so
wsrep_sst_method=xtrabackup-v2
wsrep_sst_auth="sst:defaultchannel"
wsrep_node_address=10.126.160.3
wsrep_slave_threads=8 [mysqldump]
quick max_allowed_packet = 1M [mysql]
no_auto_rehash # Only allow UPDATEs and DELETEs that use keys.
#safe_updates [myisamchk]
key_buffer_size = 512M
sort_buffer_size = 512M
read_buffer = 8M
write_buffer = 8M [mysqlhotcopy]
interactive_timeout [mysqld_safe]
open_files_limit = 65535
（c）初始化 mysql
sudo ./bin/mysqld --defaults-file=/home/worker/XtraDB_Cluster/XtraDB_3307/my.cnf --basedir=/home/worker/XtraDB_Cluster/XtraDB_3307 --initialize-insecure --user=worker
（d）启动mysql 并 加入cluster
nohup ./bin/mysqld --defaults-file=/home/worker/XtraDB_Cluster/XtraDB_3307/my.cnf &
（f）检查mysql 的状态和cluster size
（4）按照普通节点启动的方式重启第一个节点
至此，PXC 搭建完毕
搭建高可用mysql系列（2）-- Percona XtraDB Cluster 安装的更多相关文章
- MySQL高可用方案－PXC(Percona XtraDB Cluster)环境部署详解
MySQL高可用方案-PXC(Percona XtraDB Cluster)环境部署详解 Percona XtraDB Cluster简称PXC.Percona Xtradb Cluster的实现是在 ...
- 搭建高可用mysql系列（1）-- Percona XtraDB Cluster介绍
Percona XtraDB Cluster (下文简称PXC)是一个开源的mysql 高可用解决方案.它将Percona Server和Percona XtraBackup与Galera库集成在一起 ...
- mysql高可用之PXC（Percona XtraDB Cluster）
简介 Percona XtraDB Cluster是MySQL高可用性和可扩展性的解决方案,Percona XtraDB Cluster提供的特性如下: 1).同步复制,事务要么在所有节点提交或不提交 ...
- MySQL集群Percona XtraDB Cluster安装搭建步骤详解
http://www.linuxidc.com/Linux/2017-05/143501.htm http://blog.csdn.net/thundermeng/article/details/52 ...
- Percona XtraDB Cluster
简介 Percona XtraDB Cluster是MySQL高可用性和可扩展性的解决方案,Percona XtraDB Cluster提供的特性如下: 1.同步复制,事务要么在所有节点提交或不提交. ...
- Galera Cluster 实现mysql的高可用 (Percona XtraDB Cluster)
Galera Cluster 实现mysql的高可用 (Percona XtraDB Cluster) # 基础搭建 # 设备:三台主机 192.168.47.101 192.168.47.102 1 ...
- 【Linux】【Database】【MySQL】使用percona搭建高可用的MySQL数据库
1. 简介 1.1. 官方文档: 数据库架构:https://docs.openstack.org/ha-guide/shared-database.html 1.2. 本次使用的的是Percona ...
- 15、高可用 PXC（percona xtradb cluster） 搭建
安装环境: 集群名 pxc_lk 节点1: 192.168.1.20 节点2: 192.168.1.21 节点3: 192.168.1.22 所有节点安装 wget http://www.perc ...
- Percona XtraDB Cluster高可用与状态快照传输(PXC 5.7 )
Percona XtraDB Cluster(下称PXC)高可用集群支持任意节点在运行期间的重启,升级或者意外宕机,即它解决了单点故障问题.那在这个意外宕机或者重启期间,该节点丢失的数据如何再次进行同 ...
随机推荐
- 【转】详解spring 每个jar的作用
spring.jar 是包含有完整发布模块的单个jar 包.但是不包括mock.jar, aspects.jar, spring-portlet.jar, and spring-hibernate2. ...
- Spring系列之——使用了哪些设计模式
1 工厂模式:BeanFactory.ApplicationContext创建中 2 模板模式:BeanFactory.ApplicationContext实现中 3 代理模式:在AOP实现中用到了J ...
- Tomcat启动慢原因之二 he APR based Apache Tomcat Native library which allows optimal performance in production environments was not found on the java.library.path
Tomcat启动时提示: 信息: The APR based Apache Tomcat Native library which allows optimal performance in prod ...
- hihocoder 1485----hiho字符串
hihocoder 1485:hiho字符串 描述 如果一个字符串恰好包含2个'h'.1个'i'和1个'o',我们就称这个字符串是hiho字符串. 例如"oihateher".&q ...
- CentOS7系列--1.4CentOS7服务
CentOS7服务管理 1. 查看服务 1.1. 查看所有运行的服务 [root@centos7 ~]# systemctl -t service UNIT LOAD ACTIVE SUB DESCR ...
- JS实现填报时对修改过的单元格进行标识
1. 描述 在填报预览时,对单元格编辑后,其左上角有个红色标记,但非常不明显,用户很难注意到.有没有什么好的办法,对单元格操作后,将其做较明显的特殊标记处理,方便用户识别呢? 如图所示:对单元格进行操 ...
- RNN & LSTM & GRU 的原理与区别
RNN 循环神经网络,是非线性动态系统,将序列映射到序列,主要参数有五个:[Whv,Whh,Woh,bh,bo,h0][Whv,Whh,Woh,bh,bo,h0],典型的结构图如下: 和普通神经网 ...
- 灰度图的直方图均衡化(Histogram Equalization)原理与 Python 实现
原理 直方图均衡化是一种通过使用图像直方图,调整对比度的图像处理方法:通过对图像的强度(intensity)进行某种非线性变换,使得变换后的图像直方图为近似均匀分布,从而,达到提高图像对比度和增强图片 ...
- 使用 Azure CLI 创建和管理 Linux VM
Azure 虚拟机提供完全可配置的灵活计算环境. 本教程介绍 Azure 虚拟机的基本部署项目,例如选择 VM 大小.选择 VM 映像和部署 VM. 你将学习如何执行以下操作: 创建并连接到 VM 选 ...
- SQL SERVER常用语法汇总
阅读目录 一.SQL分类 二.基础语句 三.sql技巧 四.(MS SQL Server)SQL语句导入导出大全 回到目录 一.SQL分类 DDL—数据定义语言(CREATE,ALTER,DROP,D ...