MySQL replication & cluster by Joeri de Bruin ProcoliX
May 19, 2015
MySQL replication & cluster
by Joeri de BruinProcoliX
Single vs Multiple
Multiple MySQL servers means that there are moreMySQL services, this will normaly be on more
servers.
Why multiple
1. High availability 2. On line backup
replication vs cluster
ProcesReplication1. UPDATE on master2. UPDATE written in Binlog3. Binlog is read by slave4. UPDATE on slave
Cluster1. UPDATE on MySQL2. UPDATE written to Cluster filesystem
Uses
Replication has 1 master and slavesCluster can read and write on all nodes
Replication quick howto
Procedure:1. Configure master and slave2. Stop MySQL3. Copy data tot slave4. Start MySQL on master5. Start MySQL on slave
[mysqld]server-id=2master-host=db-master.mycompany.commaster-port=3306master-user=gooeymaster-password=octopus
Cluster quick howto
Procedure:Install MySQL-MAXConfigure clusterConfigure MySQLStart MGM nodeStart ndb deamonStart MySQL
[MYSQLD]ndbclusterndb-connectstring=192.168.0.10
[MYSQL_CLUSTER]ndb-connectstring=192.168.0.10
# file "config.ini"[NDBD DEFAULT]NoOfReplicas= 1MaxNoOfOrderedIndexes=2500MaxNoOfUniqueHashIndexes=2500MaxNoOfAttributes=4000
MySQL cluster and WebGUI
MySQL cluster uses NDB cluster engine in stead ofMyIsam tables
Alter create.sql to use with MySQL cluster
vi create.sql:0,$s/TYPE=MyISAM/ENGINE=NDBCLUSTER/g
Tips and tricks
Always use the same MySQL versionUse the source: www.mysql.com
Replication:http://dev.mysql.com/doc/mysql/en/replication.html
Cluster:http://dev.mysql.com/doc/mysql/en/multi-computer.html