PXC是Percona XtraDB Cluster的缩写，是一种具有高可用性和高扩展性的MySQL开源集群。它集成了Percona Server和Percona XtraBackup，同时采用了Codership Galera库。
优点：
1.准同步复制
2.多个可同时读写节点，可实现写扩展，较分片方案更进一步
3.自动节点管理
4.数据严格一致
5.服务高可用
缺点：
1.只支持innodb引擎
2.所有表都要有主键
3.所有的写操作都将发生在所有节点上，存在写扩大问题
4.加入新节点，开销大。需要复制完整的数据。
部署准备工作
| 主机名 | IP |
| pxc-linux-28 | 192.168.253.28 |
| pxc-linux-29 | 192.168.253.29 |
| pxc-linux-30 | 192.168.253.30 |
| haproxy-linux-31 | 192.168.253.31 |
三个PXC节点分别配置防火墙
[[email protected] ~]# firewall-cmd --add-port=3306/tcp --permanent [[email protected] ~]# firewall-cmd --add-port=4567/tcp --permanent [[email protected] ~]# firewall-cmd --add-port=4568/tcp --permanent [[email protected] ~]# firewall-cmd --add-port=4444/tcp --permanent [[email protected] ~]# firewall-cmd --reload
安装相关yum源
rpm -Uvh https://www.percona.com/downloads/percona-release/redhat/latest/percona-release-0.1-4.noarch.rpm
安装pxc
yum install Percona-XtraDB-Cluster-57 -y
分别在三个节点创建数据存储目录
mkdir -p /data/local/percona-xtradb-cluster/data chown -R mysql:mysql /data/local/percona-xtradb-cluster/data mkdir -p /data/local/percona-xtradb-cluster/run chown -R mysql:mysql /data/local/percona-xtradb-cluster/run mkdir -p /data/logs/mysql chown -R mysql:mysql /data/logs/mysql
修改/etc/my.cnf配置文件，其他两台节点需要修改server_id和wsrep_node_address
pxc-linux-28:
#cat /etc/my.cnf [client] port = 3306 socket = /data/local/percona-xtradb-cluster/run/mysql.sock default-character-set = utf8mb4 [mysqld] user = mysql innodb_buffer_pool_size = 1024M character_set_server = utf8mb4 datadir = /data/local/percona-xtradb-cluster/data port = 3306 server_id = 28 socket = /data/local/percona-xtradb-cluster/run/mysql.sock pid-file = /data/local/percona-xtradb-cluster/run/mysql.pid log-error = /data/logs/mysql/error.log log_warnings = 2 slow_query_log_file = /data/logs/mysql/slow.log long_query_time = 2 log_timestamps=SYSTEM lower_case_table_names = 1 key_buffer_size = 1344M event_scheduler=ON sql_mode=NO_ENGINE_SUBSTITUTION,STRICT_TRANS_TABLES wsrep_provider=/usr/lib64/libgalera_smm.so wsrep_cluster_address=gcomm://192.168.253.28,192.168.253.29,192.168.253.30 binlog_format=ROW pxc_strict_mode=PERMISSIVE max_connect_errors=1000 max_allowed_packet = 1024M default_storage_engine=InnoDB #Innodb innodb_flush_method = O_DIRECT innodb_log_files_in_group = 5 innodb_lock_wait_timeout = 50 innodb_log_file_size = 1024M innodb_flush_log_at_trx_commit = 1 innodb_file_per_table = 1 innodb_thread_concurrency = 8 innodb_buffer_pool_size = 5G innodb_read_io_threads = 24 innodb_write_io_threads = 24 log_bin_trust_function_creators=1 innodb_autoinc_lock_mode=2 innodb_locks_unsafe_for_binlog=1 # CACHES AND LIMITS # tmp_table_size = 32M max_heap_table_size = 32M max_connections = 1000 thread_cache_size = 50 open_files_limit = 65535 table_definition_cache = 4096 table_open_cache = 5000 #wsrep wsrep_retry_autocommit=1 wsrep_auto_increment_control=1 wsrep_node_name = pxc-linux-28 wsrep_node_address=192.168.253.28 wsrep_sst_method=xtrabackup-v2 wsrep_cluster_name=test-pxc wsrep_sst_auth="sync_rep:abc123"
pxc-linux-29:
[client] port = 3306 socket = /data/local/percona-xtradb-cluster/run/mysql.sock default-character-set = utf8mb4 [mysqld] user = mysql innodb_buffer_pool_size = 1024M character_set_server = utf8mb4 datadir = /data/local/percona-xtradb-cluster/data port = 3306 server_id = 29 socket = /data/local/percona-xtradb-cluster/run/mysql.sock pid-file = /data/local/percona-xtradb-cluster/run/mysql.pid log-error = /data/logs/mysql/error.log log_warnings = 2 slow_query_log_file = /data/logs/mysql/slow.log long_query_time = 2 log_timestamps=SYSTEM log_bin_trust_function_creators=1 lower_case_table_names=1 key_buffer_size = 1344M event_scheduler=ON sql_mode=NO_ENGINE_SUBSTITUTION,STRICT_TRANS_TABLES wsrep_provider=/usr/lib64/libgalera_smm.so wsrep_cluster_address=gcomm://192.168.253.28,192.168.253.29,192.168.253.30 binlog_format=ROW pxc_strict_mode=PERMISSIVE max_connect_errors=1000 max_allowed_packet = 1024M default_storage_engine=InnoDB #Innodb innodb_flush_method = O_DIRECT innodb_log_files_in_group = 5 innodb_lock_wait_timeout = 50 innodb_log_file_size = 1024M innodb_flush_log_at_trx_commit = 1 innodb_file_per_table = 1 innodb_thread_concurrency = 8 innodb_buffer_pool_size = 5G innodb_read_io_threads = 24 innodb_write_io_threads = 24 log_bin_trust_function_creators=1 innodb_autoinc_lock_mode=2 innodb_locks_unsafe_for_binlog=1 # CACHES AND LIMITS # tmp_table_size = 32M max_heap_table_size = 32M max_connections = 1000 thread_cache_size = 50 open_files_limit = 65535 table_definition_cache = 4096 table_open_cache = 5000 #wsrep wsrep_retry_autocommit=1 wsrep_auto_increment_control=1 wsrep_node_name =pxc-linux-29 wsrep_node_address=192.168.253.29 wsrep_sst_method=xtrabackup-v2 wsrep_cluster_name=test-pxc wsrep_sst_auth="sync_rep:abc123"
pxc-linux-30:
[client] port = 3306 socket = /data/local/percona-xtradb-cluster/run/mysql.sock default-character-set = utf8mb4 [mysqld] user = mysql innodb_buffer_pool_size = 1024M character_set_server = utf8mb4 datadir = /data/local/percona-xtradb-cluster/data port = 3306 server_id = 30 socket = /data/local/percona-xtradb-cluster/run/mysql.sock pid-file = /data/local/percona-xtradb-cluster/run/mysql.pid log-error = /data/logs/mysql/error.log log_warnings = 2 slow_query_log_file = /data/logs/mysql/slow.log long_query_time = 2 log_timestamps=SYSTEM lower_case_table_names = 1 key_buffer_size = 1344M event_scheduler=ON sql_mode=NO_ENGINE_SUBSTITUTION,STRICT_TRANS_TABLES wsrep_provider=/usr/lib64/libgalera_smm.so wsrep_cluster_address=gcomm://192.168.253.28,192.168.253.29,192.168.253.30 binlog_format=ROW pxc_strict_mode=PERMISSIVE max_connect_errors=1000 max_allowed_packet = 1024M default_storage_engine=InnoDB #Innodb innodb_flush_method = O_DIRECT innodb_log_files_in_group = 5 innodb_lock_wait_timeout = 50 innodb_log_file_size = 1024M innodb_flush_log_at_trx_commit = 1 innodb_file_per_table = 1 innodb_thread_concurrency = 8 innodb_buffer_pool_size = 5G innodb_read_io_threads = 24 innodb_write_io_threads = 24 log_bin_trust_function_creators=1 innodb_autoinc_lock_mode=2 innodb_locks_unsafe_for_binlog=1 # CACHES AND LIMITS # tmp_table_size = 32M max_heap_table_size = 32M max_connections = 1000 thread_cache_size = 50 open_files_limit = 65535 table_definition_cache = 4096 table_open_cache = 5000 #wsrep wsrep_retry_autocommit=1 wsrep_auto_increment_control=1 wsrep_node_name = pxc-linux-30 wsrep_node_address=192.168.253.30 wsrep_sst_method=xtrabackup-v2 wsrep_cluster_name=test-pxc wsrep_sst_auth="sync_rep:abc123"
pxc-linux-28启动
systemctl start [email protected]
备注：谨记，只要是启动集群的第一个Node（首次搭建集群或者集群全部关闭），都要用此命令
pxc-linux-29和pxc-linux-30启动
systemctl start mysql
备注：谨记，只要集群有一个Node启动，其余节点都是用此命令
修改密码
mysql5.7版本日志均在error.log 里面生成
grep "temporary password" /data/logs/mysql/error.log
使用改密码登陆MySQL，修改成自己想要的密码
mysql> alter user 'root'@'localhost' idnetified by '123456';
在pxc-linux-28MySQL中创建SST同步用户
mysql> GRANT PROCESS,RELOAD,LOCK TABLES,REPLICATION CLIENT ON *.* TO 'sync_rep'@'192.168.%.%' IDENTIFIED BY 'abc123'; mysql> flush privileges;
State Snapshot Transfer 是可以让one node (donor) 通过数据拷贝升级至joining node (joiner)。
一个新节点加入至集群中，需要同步数据，那么SST可以传输数据，让新节点成为集群的一部分。
SST的三个有效方法：mysqldump, rsync andxtrabackup。不过多解释，目前最好用的是xtrabackup
查看集群状态
集群验证
[[email protected] ~]# mysql -uroot -p mysql> show variables like 'version'; +---------------+--------------+ | Variable_name | Value | +---------------+--------------+ | version | 5.7.19-17-57 | +---------------+--------------+ 1 row in set (0.00 sec) mysql> create database pxcdb; mysql> use pxcdb; mysql> create table t1(id tinyint,ename varchar(20)); mysql> insert into t1 values(1,'Leshami'); 在29上进行验证 [[email protected] ~]# mysql -uroot -p mysql> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | pxcdb | | sys | +--------------------+ 5 rows in set (0.00 sec) mysql> show variables like 'server_id'; +---------------+-------+ | Variable_name | Value | +---------------+-------+ | server_id | 29 | +---------------+-------+ 1 row in set (0.01 sec) mysql> select * from pxcdb.t1; +------+---------+ | id | ename | +------+---------+ | 1 | Leshami | +------+---------+ 1 row in set (0.00 sec) --查看集群状态 mysql> show status like '%wsrep_clust%'; +--------------------------+--------------------------------------+ | Variable_name | Value | +--------------------------+--------------------------------------+ | wsrep_cluster_conf_id | 13 | | wsrep_cluster_size | 3 | | wsrep_cluster_state_uuid | aeb87793-ebb2-11e7-b33e-eeaf4988bbe4 | | wsrep_cluster_status | Primary | +--------------------------+--------------------------------------+ 4 rows in set (0.00 sec) mysql> show status like 'wsrep_connected'; +-----------------+-------+ | Variable_name | Value | +-----------------+-------+ | wsrep_connected | ON | +-----------------+-------+ 1 row in set (0.00 sec)
openssl版本过低导致的错误
2017-12-28T09:23:19.605353Z 0 [ERROR] WSREP: wsrep_load(): dlopen(): /usr/lib64/galera3/libgalera_smm.so: symbol SSL_COMP_free_compression_methods, version libssl.so.10 not defined in file libssl.so.10 with link time reference 2017-12-28T09:23:19.605379Z 0 [ERROR] WSREP: Failed to load wsrep_provider (/usr/lib64/galera3/libgalera_smm.so). Error: Invalid argument (code: 22). Reverting to no provider. 2017-12-28T09:23:19.605386Z 0 [Note] WSREP: Setting wsrep_ready to false [[email protected] ~]# rpm -qa|grep openssl openssl-1.0.1e-42.el7.9.x86_64 openssl-libs-1.0.1e-42.el7.9.x86_64 [[email protected] ~]# yum update openssl -y
需要升级openssl，建议全部升级后再启动集群，这问题在CentOS 7.4.1708不存在即openssl版本较新
Haproxy负载均衡
Haproxy是一个反向代理负载均衡解决方案，支持4层和7层模式，提供后端服务器健康检查，非常稳定。淘宝前期也使用Haproxy作为CDN系统负载均衡器。
安装haproxy
yum -y install haproxy
简单配置文件：
cat /etc/haproxy/haproxy.cfg #--------------------------------------------------------------------- # Example configuration for a possible web application. See the # full configuration options online. # # http://haproxy.1wt.eu/download/1.4/doc/configuration.txt # #--------------------------------------------------------------------- #--------------------------------------------------------------------- # Global settings #--------------------------------------------------------------------- global # to have these messages end up in /var/log/haproxy.log you will # need to: # # 1) configure syslog to accept network log events. This is done # by adding the '-r' option to the SYSLOGD_OPTIONS in # /etc/sysconfig/syslog # # 2) configure local2 events to go to the /var/log/haproxy.log # file. A line like the following can be added to # /etc/sysconfig/syslog # # local2.* /var/log/haproxy.log # log 127.0.0.1 local2 chroot /var/lib/haproxy pidfile /var/run/haproxy.pid maxconn 4000 user haproxy group haproxy daemon # turn on stats unix socket stats socket /var/lib/haproxy/stats #--------------------------------------------------------------------- # common defaults that all the 'listen' and 'backend' sections will # use if not designated in their block #--------------------------------------------------------------------- defaults mode http log global option httplog option dontlognull option http-server-close option forwardfor except 127.0.0.0/8 option redispatch retries 3 timeout http-request 10s timeout queue 1m timeout connect 10s timeout client 1m timeout server 1m timeout http-keep-alive 10s timeout check 10s maxconn 3000 #--------------------------------------------------------------------- # main frontend which proxys to the backends #--------------------------------------------------------------------- frontend pxc-front bind *:3306 mode tcp default_backend pxc-back frontend statas-front bind *:8088 mode http default_backend stats-back #--------------------------------------------------------------------- # static backend for serving up images, stylesheets and such #--------------------------------------------------------------------- backend pxc-back mode tcp balance leastconn option httpchk server pxc-linux-28 192.168.253.28:3306 check port 9200 inter 12000 rise 3 fall 3 server pxc-linux-29 192.168.253.29:3306 check port 9200 inter 12000 rise 3 fall 3 server pxc-linux-30 192.168.253.30:3306 check port 9200 inter 12000 rise 3 fall 3 #--------------------------------------------------------------------- # round robin balancing between the various backends #--------------------------------------------------------------------- backend stats-back mode http balance roundrobin stats uri /haproxy/stats stats auth pxcstats:xxwzopop
在cluster的MySQL上创建用户
(一个节点创建，会被复制到其它节点)
监控用帐号：
grant usage on *.* to 'pxc-monitor'@'%' identified by 'testpxc';
服务测试帐号：
grant all privileges on *.* to 'zxw'@'%' identified by 'xxwzopop';
更改用户名和密码（三个节点都得修改）
#vim /usr/bin/clustercheck MYSQL_USERNAME="pxc-monitor" MYSQL_PASSWORD="testpxc"
测试
[[email protected] ~]# clustercheck HTTP/1.1 200 OK Content-Type: text/plain Connection: close Content-Length: 40 Percona XtraDB Cluster Node is synced.
更改/etc/services添加mysqlchk的端口号：
mysqlchk 9200/tcp #mysqlchk
配置xinetd
vi /etc/xinetd.d/mysqlchk
该文件保持默认即可
测试
#for i in `seq 1 1000`;do mysql -h 192.168.253.31 -P3306 -upxc -pabc123 -e "select @@hostname;";done mysql: [Warning] Using a password on the command line interface can be insecure. +--------------+ | @@hostname | +--------------+ | pxc-linux-30 | +--------------+ mysql: [Warning] Using a password on the command line interface can be insecure. +--------------+ | @@hostname | +--------------+ | pxc-linux-28 | +--------------+ mysql: [Warning] Using a password on the command line interface can be insecure. +--------------+ | @@hostname | +--------------+ | pxc-linux-29 | +--------------+
可以看到负载了
查看haproxy状态
http://192.168.253.31:8088/haproxy/stats
帐号密码是：pxcstats:xxwzopop
/etc/haproxy/haproxy.cfg 指定的。
原文地址：http://blog.51cto.com/11243465/2104885
时间： 2024-08-06 05:05:05