MariaDB-yum仓库搭建方法：
相关参考网站：
http://yum.mariadb.org/
https://mariadb.com/kb/en/mariadb/yum/
MariaDB 10.X
centos6 repo
# MariaDB 10.1 CentOS repository list - created 2015-11-20 14:01 UTC
# http://mariadb.org/mariadb/repositories/
[mariadb]
name = MariaDB
baseurl = http://yum.mariadb.org/10.1/centos6-amd64
gpgkey=https://yum.mariadb.org/RPM-GPG-KEY-MariaDB
gpgcheck=1
centos7 repo
[mariadb]
name = MariaDB
baseurl = http://yum.mariadb.org/10.1/centos7-amd64
gpgkey=https://yum.mariadb.org/RPM-GPG-KEY-MariaDB
gpgcheck=1
MariaDB 5.X
centos6 repo
[mariadb]
name = MariaDB-5.5.39
baseurl=https://downloads.mariadb.com/files/MariaDB/mariadb-5.5.39/yum/rhel6-amd64/
# alternative: baseurl=http://archive.mariadb.org/mariadb-5.5.39/yum/rhel6-amd64/
gpgkey=https://yum.mariadb.org/RPM-GPG-KEY-MariaDB
gpgcheck=1
安装MariaDB 10.1
yum install MariaDB-server MariaDB-client -y
初始化数据库：
/usr/bin/mysql_install_db --datadir=/data/mysql --user=mysql
配置参数：
vim /etc/my.cnf.d/server.cnf
[mysqld]
server-id = 1
datadir = /data/mysql
log_error = /data/mysql/err.log
log_bin = mysql-bin
slow_query_log = ON
slow_query_log_file = /data/mysql/slow.log
long_query_time = 2
skip-name-resolve
lower_case_table_names = 1
innodb_file_per_table = 1
back_log = 50
max_connections = 2000
max_connect_errors = 1000000
thread_cache_size = 128
table_open_cache = 2048
max_allowed_packet = 1G
binlog_cache_size = 8M
character_set_server = utf8
innodb_data_file_path = ibdata1:10M;ibdata2:10M:autoextend
innodb_thread_concurrency = 0
innodb_read_io_threads = 10
innodb_write_io_threads = 10
innodb_flush_method=O_DIRECT
innodb_buffer_pool_size = 400M
innodb_log_file_size = 100M
innodb_log_buffer_size = 16M
innodb_log_files_in_group = 3
innodb_max_dirty_pages_pct = 60
innodb_lock_wait_timeout = 60
log_bin_trust_function_creators = 1
innodb_flush_log_at_trx_commit = 2
innodb_locks_unsafe_for_binlog = 1
#innodb_log_group_home_dir =
[galera]
# Mandatory settings
wsrep_on=ON
#wsrep_provider=
wsrep_provider = /usr/lib64/galera/libgalera_smm.so
wsrep_cluster_address= 'gcomm://192.168.0.241:4567,192.168.0.243:4567'
binlog_format=row
default_storage_engine=InnoDB
innodb_autoinc_lock_mode=2
wsrep_retry_autocommit = 0
wsrep_provider_options="gcache.size=200M; base_port=4567; gcache.page_size=100M"
wsrep_slave_threads=4
wsrep_cluster_name='mysql_db_cluster'
wsrep_node_address='192.168.0.242'
wsrep_provider_options="base_port=4567;"
wsrep_node_name='test2'
wsrep_sst_method=rsync
wsrep_sst_auth="sst:86rt6y1qaz@WSX"
# Allow server to accept connections on all interfaces.
bind-address=0.0.0.0
启动第一个数据库：192.168.0.241
/etc/init.d/mysql start --wsrep-new-cluster
注：在集群中启动第一台数据库时，必须加上参数--wsrep-new-cluster
依次启动多台数据库：192.168.0.242 192.168.0.243
/etc/init.d/mysql start --wsrep_cluster_address=gcomm://192.168.0.241
注：依次启动多台数据库时，需要加上参数--wsrep_cluster_address并指定IP地址，以加入集群
当集群中的所有数据库全部关闭后第一台启动的数据库：
需要添加 --wsrep-new-cluster选择
当集群中只要有一台数据库未关闭，其它集群节点启动不需要添加任何额外的参数
注：关闭数据库一定要正常关闭，否则数据库可能会无法启动
查看集群状态：
mysql>show status like ‘wsrep%’;
初次启动碰到的故障：
[ERROR] InnoDB: auto-extending data file ./ibdata1 is of a different size 640 pages (rounded down to MB) than specified in the .cnf file: initial 768 pages, max 0 (relevant if non-zero) pages!
问题原因：
在初始化数据库时系统自动建立了ibdata1,ib_logfile0,ib_logfile1几个文件
在配置文件中由于配置了：innodb_data_file_path = ibdata1:10M;ibdata2:10M:autoextend，跟初始化时的ibdata1的大小不一致导致的
默认的ibdata1大小为8M，而我在此配置的为10M，是导致问题的原因
解决办法：
在mysql的data目录下，删除ibdata1,ib_logfile0,ib_logfile1；
在重新启动mysql，系统会根据配置再次生成ibdata文件
