HIGH AVAILABILITY MYSQL DATABASE REPLICATION WITH SOLARIS ZONE CLUSTER

Pedro Lay, Technical Systems Marketing Sun BluePrints Online

Part No 820-7582-10 Revision 1.0, 2/23/09

Sun Microsystems, Inc.

Table of Contents
MySQL database replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Solaris Zones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Solaris Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Example configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Failover scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 About the author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Ordering Sun documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Accessing Sun documentation online . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Appendix A: Command files for clzonecluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Appendix B: MySQL configuration file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Appendix C: Configuration files: /config-files/mysql_config . . . . . . . . . . . . . . . . . 17 Appendix D: Configuration files: /config-files/ha_mysql_config . . . . . . . . . . . . . . 19

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

High Availability MySQL Database Replication with Solaris Zone Cluster

New technology offers new possibilities to accomplish things in a different and more efcient way, moreover when there is a convergence of three new technologies that complements and provides additional capabilities to solve business requirements. Such is the case when Sun MySQL Replication, Solaris Containers and Solaris Zone Cluster come into play. MySQL Replication is an option which allows the content of one database to be replicated to another database or databases, providing a mechanism to scale out the database. Scaling out the database allows more activities to be processed and more users to access the database by running multiple copies of the databases on different machines. Solaris Containers provide a virtualized runtime environment by using Solaris Zones partitioning technology and resource management tools that are part of the Solaris operating system; the zone is a container for an application. With Solaris Containers, MySQL Replication can be run with the databases replicated across zones. The latest release of Solaris Cluster 3.2 1/09 introduces the concept of a zone cluster consisting of a set of virtual nodes where each virtual node is a Solaris Zone. This new feature of zone clustering, along with the High Availability service for MySQL provided by Solaris Cluster, allows automatic failover across machines of a MySQL Replication deployed with Solaris Zones. This paper describes the benet of deploying the master and slave database of MySQL Replication using zone clusters. In addition, it provides details on how to deploy the master and slave databases in two different zone clusters, using non-global zones from two different machines as the virtual cluster nodes.

MySQL database replication

Replication is an option that enables data changes from one MySQL database server, called the master, to be duplicated to one or more MySQL database servers, called slaves. The replication of data is performed in a one way asynchronous mode. The replication is one way because data is only updated at the master database while data retrieval is the only operation that can be performed at the slave databases. It is asynchronous because the master database does not need to wait for the data to be actually replicated in the slave databases before continuing its operation. This type of replication is a poll model, where the slave is connected to a master and asks the master to send events the slave does not have.

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

Depending on how replication is congured, an entire database or selected tables within a database can be replicated. And, depending on how replication is deployed to fulll a business need, MySQL Replication can be deployed with the following topologies:

Master

Slave

Master

Slaves

Master

Slave

Slaves

Figure 1. MySQL database replication topologies.

Note More complex topologies are possible and have been deployedsuch as mastermaster replications, where both servers act as both master and slave. While these congurations are supported, care must be taken to ensure that they are congured correctly to avoid overwriting of data.

A number of threads and les from both the master and slave database servers are involved when replication is enabled, as shown in Figure 2.
Master Server Slave Server

Database

IO Thread

SQL Thread

Database

Binary Log

Relay Log

.info files

Figure 2. MySQL replication process. From the master database server: Binary Log lewhere updates of the master database are captured IO_Threadcaptures the updates of the master to the Binary Log le From the slave database server: Relay Log lecopy of the binary log from master database Relay Index logindex of the Relay Log le Master.info lecontains all information about the master server Relay-log.info lecontains information of the SQL_Thread IO_Threaddownloads the Binary Log le from the master into the Relay Log le SQL_Threadexecutes SQL commands from the Relay Log le

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

Solaris Zones
Sun provides many technologies for hosting multiple applications on a single system hardware, including hardware partitions (Dynamic Domains), virtual machines (Sun Logical Domains, Sun xVM software), operating system virtualization (Solaris Zones) and resource management. The Solaris OS provides zones that enable software partitioning of a single Solaris 10 OS instance to support multiple virtual independent operating systems with independent process space, allocated resources, and users. There are two types of zones: global and non-global. Every Solaris system contains a global zone: it is the default zone for the system, and system infrastructure administration (such as conguring physical devices, routing, etc.) occurs in the global zone. Non-global zones contain an installed subset of the complete Solaris operating system software packages and provide security isolation, application fault isolation and a resource-managed environment to run applications. Solaris zones allow both the master database and the slave database of MySQL Replication to run independently and isolated from each other on the same machine.

Solaris Cluster
A cluster is two or more systems, or nodes, that work together as a single entity to provide increased availability and/or performance. Solaris Cluster provides failover services for MySQL database by making the database able to survive any single software or hardware failure in the system. When a failure occurs, the MySQL database is restarted on the surviving node of the cluster without user intervention. MySQL database can be deployed in the following conguration on Solaris Clusters: MySQL database in the global zone MySQL database in the nonglobal failover zone MySQL database in a non-global zone MySQL database in a zone cluster The Sun BluePrints Online article Deploying MySQL Database in Solaris Cluster Environments for Increased High Availability provides a good overview of Solaris Cluster and its components, and provides an example on deploying MySQL databases in the global zone. The release of Solaris Cluster 3.2 1/09 introduces the concept of a zone cluster consisting of a set of virtual nodes, where each node is a Solaris zone. The zone cluster extends the Solaris zone principles to work across a cluster. The following section describes how the master and slave database of MySQL Replication were deployed in two different zone clusters, using the cluster-branded non-global zones from two different machines as the virtual cluster nodes.

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

Example conguration
The following example conguration was used to deploy MySQL database replication using Solaris Zone Clusters. Hardware conguration: Two servers (a Sun SPARC Enterprise T5220 and T5140 server) with the following components: Local disk with a 512 MB slice for /globaldevices Three network ports: one port to be used for the public network interface, and two ports to be used for the Cluster Interconnect (private network) Two storage interface ports Multihost storage device for the shared disks (Sun StorEdge 3510 FC Array) Software conguration: Solaris 10 5/08 s10s_u5wos_10 SPARC System Firmware (sysfw_version): 7.1.6 2008/08/15 02:51 SP Firmware Version/Build Nbr: 2.0.4.26 / 35945 Solaris Cluster 3.2 1/09 MySQL 5.0.22 Patches for MySQL Data Services: 126032-05 for Solaris 10 OS (SPARC) 126031-05 for Solaris 9 OS (SPARC) 126033-06 for Solaris 10 OS (x86) The actual hardware conguration used for deployment is shown in Figure 3.

Sun SPARC Enterprise T5220 Server

0 1 2 3

0 1 2 3

Sun SPARC Enterprise T5140 Server

PCI 0

P1

P2

P1

P2

PCI 3

FC 0

FC 1

FC 2

FC 3

Local Disks

3 LUNS of 20GB each

Sun StorEdge 3510 FC Array

Local Disks

Figure 3. Example hardware conguration.

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

The following steps were followed to install and congure Solaris Cluster and MySQL Data Services: 1. 2. 3. Install Solaris Cluster software on both servers, then run the installer command to invoke the Solaris Cluster installer GUI. Congure the Solaris Cluster software from one node of the cluster by executing the command /usr/cluster/bin/scinstall. Solaris Zone Cluster Conguration: create two zone clusters (sparse root) named ZC2 and ZC3 with the following characteristics: Two cluster nodes using non-global zones from two different machines. A storage volume accessible by both cluster nodes. This is the High Availability Storage Plus (HASP) resource. A public IP address. This is the Logical Host (LH) resource. A logical diagram of the Solaris zone conguration is depicted in Figure 4.

SC Node: tm16-180
Zone: ZC2 Logical Host: Ih-2

SC Node: tm16-184

Zone Host: paris

Zone Host: mumbai

Zone: ZC3 Logical Host: Ih-3

Zone Host: rome

Zone Host: bangalore

/pool3

/pool2

Figure 4. Example Solaris zone conguration. Zone Cluster ZC2: Used for the master database Logical Host name: lh-2 Zone host names: paris and mumbai Has Logical Host resource and HASP resource on a zpool HA ZFS file system /pool2 on shared storage

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

Zone Cluster ZC3: Used for the slave database Logical Host name: lh-3 Zone host names: rome and bangalore Has Logical Host resource and HASP resource on a zpool HA ZFS file system /pool3 on shared storage

The command les used to congure Zone Cluster ZC2 and ZC3 are listed in Appendix A, Command les for clzonecluster on page 13. The creation of the zone cluster is done by executing the following commands:
# clzonecluster configure -f cmd-file-2 zc2 # clzonecluster configure -f cmd-file-3 zc3

4.

The following commands were used to congure Solaris Cluster Resource for MySQL Data Service. These commands perform the following tasks: Create failover resource group and logical host name. Register the HAStoragePlus resource type. Create a HAStorage resource to the resource group for the local ZFS. Bring the resource group online.

In MySQL1 zone 2:
paris @ / $ clrg create mysql-1-rg paris @ / $ clreslogicalhostname create -g mysql-1-rg -h lh-2 mysql-1-lh paris @ / $ clrt register SUNW.HAStoragePlus paris @ / $ clrs create -g mysql-1-rg -t SUNW.HAStoragePlus -p zpools=pool2 mysql-1-hasp-rs paris @ / $ clrg online -eM mysql-1-rg

In MySQL2 zone 3:
bangalore @ / $ clrg create mysql-2-rg bangalore @ / $ clreslogicalhostname create -g mysql-2-rg -h lh-3 mysql-2-lh bangalore @ / $ clrt register SUNW.HAStoragePlus bangalore @ / $ clrs create -g mysql-2-rg -t SUNW.HAStoragePlus -p zpools=pool3 mysql-2-hasp-rs bangalore @ / $ clrg online -eM mysql-2-rg

5.

Install the MSQL Agent packages on the physical hosts: system SUNWmysqlrMySQL Database Management System (root component) system SUNWmysqltMySQL Database Management System (test component) system SUNWmysqluMySQL Database Management System (usr component)

6.

MySQL Replication deployment: A minimum of two MySQL database instances is needed to deploy MySQL Replication: The master database was installed in the first zone cluster The slave database was installed in the second zone cluster The following steps were used to setup the MySQL database for replication:

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

From the master database server: a. b. Create a new user with REPLICATION SLAVE privilege. Turn on binary log and set the MySQL database server-id to a unique number. (This requires shutting down the MySQL server and modifying the MySQL conguration le.) Determine master database position and binary log le name by executing the following MySQL commands: flush tables with read lock; show master status; Create a backup of the master database that contains binary log coordinates.

c.

d.

From the slave database server: a. b. c. d. Shutdown the database and modify the MySQL conguration le by assigning a unique number to the server-id. Load in the backup from the master database. Bring up MySQL server. Congure the slave database with master database information (such as where the master database resides and the master database position number in the binary log). This operation is performed by executing the MySQL command: CHANGE MASTER TO Start replication by executing the MySQL command: SLAVE START

e.

The conguration les of the master and slave databases are shown in Appendix B, MySQL conguration le on page 15. 7. Enable and register the MySQL database to be used under Solaris Cluster: a. Copy and edit the mysql_config le. See Appendix C, Conguration les: /cong-les/mysql_cong on page 17 for master and slave databases. Run the mysql_register script from the logical host of zone cluster ZC2 and ZC3:

b.

paris @ / $ ksh mysql_register -f /config-files/mysql_config bangalore @ / $ ksh mysql_register -f /config-files/mysql_config

c. d.

e.

Shutdown the MySQL database. Copy and edit the ha_mysql_config le. See Appendix D, Conguration les: /cong-les/ha_mysql_cong on page 19 for master and slave databases. Run the ha_mysql_register script from the logical host of zone cluster ZC2 and ZC3:

paris @ / $ ksh ha_mysql_register -f /config-files/ha_mysql_cofig bangalore@ / $ ksh ha_mysql_register -f /config-files/ha_mysql_config

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

f.

Execute the following command from the logical host of zone cluster ZC3 and ZC3:

paris @ / $ clresource enable mysql-1-hasp-rs bangalore @ / $ clresource enable mysql-2-hasp-rc

At this point, the MySQL Replication deployment is congured as shown in Figure 5.

SC Node: tm16-180
Zone Host: paris

SC Node: tm16-184
Zone Host: mumbai

Zone: ZC2 Logical Host: Ih-2 2

Master
Zone Host: Zone Host: bangalore

Zone: ZC3 Logical Host: Ih-3 3

Slave

Figure 5. Example master and slave database conguration.

Note With Solaris Cluster, sc3_test_database is created and used for monitoring purposes. This database must be excluded from replication.

Failover scenarios
Now that the MySQL Replication is deployed with Solaris Cluster, consider how an application running on a different host on the network will interact with the MySQL database, and how Solaris Cluster provides data services, with the following scenario: An application running from a host on the network connects via the logical host lh-2 to the MySQL master database to perform update transactions. A second application running from a host on the network connects via the logical host lh-3 to the MySQL slave database to do query transactions. With MySQL Replication, whatever update happens at the master database is replicated in the slave database. As a simple example, if user A is inserting rows to a table X in the master database, then user B will see the new rows of table X as he/she queries table X from the slave database.

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

The following describes the two failover scenarios: master database failover and slave database failover. Master database failover If the master database fails, Solaris Zone Cluster will fail over the master database to the surviving zone, as shown in Figure 6. During the fail over, the client application will briey pause and resume transactions after the database services are failed over. After the failover, MySQL will resume the replication of update transactions from the master database to the slave database; this time the master database will be running from the surviving node (zone host mumbai, in this example). Likewise, the client application doing queries against the slave database will briey pause and the query will resume.

SC Node: tm16-180

SC Node: tm16-184
Zone Host: mumbai

Zone: ZC2 Logical Host: Ih-2 2

X
Master Slave

Zone Host: paris

Master M
Zone Host: bangalore

Zone Host:

Zone: ZC3 Logical Host: Ih-3 3

Figure 6. Master database failover. Slave database failover If the slave database fails, Solaris Zone Cluster will fail over the slave database to the surviving zone, as shown in Figure 7. The sequence of events is similar to the master database failover scenariowith the difference that it is the slave database that is failing over. The client doing updates against the master database is not affected at all. The client doing queries against the slave database will briey pause and the query will resume after the slave database is failed over to the surviving zone.

10

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

SC Node: tm16-180
Zone Host: paris

SC Node: tm16-184
Zone Host: mumbai

Zone: ZC2 Logical Host: Ih-2 2

Master
Zone Host: Zone Host: bangalore

Zone: ZC3 Logical Host: Ih-3 3

X
Slave

Slave

Figure 7. Slave database failover. To test the failover process of both master and slave databases, this study injected fault conditions by executing zoneadm halt, clrg switch, and clrs disable commands; killing the mysql daemon; and even pulling out the power cord on one of the server machines.

Summary
The MySQL database delivers a fast, multi-threaded, multi-user, and robust SQL (Structured Query Language) database server. With the MySQL Replication option enabled, this database provides a mean to horizontally improve application performance and scalability by adding multiple replicated database servers. MySQL Replication uses a poll model, where slave databases are connected to a master database and ask the master to send events the slave does not have. There are many ways MySQL Replication can be deployed and used, such as running database backup from the slave while the master continues processing transactions, or using the master database to process updates while the slave is used for queries, to name a few. Often customers provide high availability to the master database, and Solaris Cluster HA for MySQL data service provides a mechanism for orderly startup and shutdown, fault monitoring and automatic failover of MySQL database, making recovery from failure transparent to clients. Furthermore, to improve power consumption from the multiple replicated database servers, Solaris OS provides zones to deploy MySQL Replication, and along with Solaris Cluster the zones can be part of a virtual zone.

11

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

About the author

Pedro Lay is an Enterprise Solutions Architect in Suns Systems Technical Marketing Group. He has over 20 years of industry experience that spans application development, database and system administration, and performance and tuning efforts. Since joining Sun in 1990, Pedro has worked in various organizations including Information Technology, the Customer Benchmark Center, the Business Intelligence and Data Warehouse Competency Center, and the Performance Applications Engineering group.

Acknowledgements
The author would like to recognize Gia-Khanh Nguyen, and Detlet Ulherr from the Sun Cluster group for their contributions to this article.

References
Kamboj, Ritu. Deploying MySQL Database in Solaris Cluster Environments for Increased High Availability, Sun BluePrints Online, November, 2008. http://wikis.sun.com/display/BluePrints/Deploying+MySQL+Database+in+Solar is+Cluster++Environments Kloski, Nick. MySQL Database Scale-Out and Replication for High-Growth Businesses, Sun BluePrints Online, November, 2008. http://wikis.sun.com/download/attachments/57508101/820-6824.pdf?version=1 Sun Cluster 3.2 1/09 Documentation Center: http://docs.sun.com/app/docs/doc/820-4683?l=en Sun Cluster 3.2 1/09 Release Notes http://wikis.sun.com/display/SunCluster/(English)+Sun+Cluster+3.2+109+Release+Notes#(English)SunCluster3.21-09ReleaseNotes-zc Sun Cluster Data Service for MySQL Guide for Solaris OS (January 2009) http://docs.sun.com/app/docs/doc/820-5027?l=en

Ordering Sun documents

The SunDocsSM program provides more than 250 manuals from Sun Microsystems, Inc. If you live in the United States, Canada, Europe, or Japan, you can purchase documentation sets or individual manuals through this program.

12

High Availability MySQL Database Replication with Solaris Zone Clus-

Sun Microsystems, Inc.

Accessing Sun documentation online

The docs.sun.com web site enables you to access Sun technical documentation online. You can browse the docs.sun.com archive or search for a specic book title or subject. The URL is http://docs.sun.com/ To reference Sun BluePrints Online articles, visit the Sun BluePrints Online Web site at: http://www.sun.com/blueprints/online.html

13

Command les for clzonecluster

Sun Microsystems, Inc.

Appendix A

Command les for clzonecluster

The clzonecluster command le used for Zone Cluster zc2:
create set zonepath=/zones/zc2 add sysid set root_password=ZiitH.NOLOrRg set name_service=NONE set nfs4_domain=dynamic set security_policy=NONE set system_locale=C set terminal=xterms set timezone=US/Pacific end add node set physical-host=tm16-180 set hostname=paris add net set address=paris/24 set physical=nxge0 end end add node set physical-host=tm16-184 set hostname=mumbai add net set address=mumbai/24 set physical=e1000g0 end end add net set address=lh-2 end commit

14

Command les for clzonecluster

Sun Microsystems, Inc.

The clzonecluster command le used for Zone Cluster zc3:

create set zonepath=/zones/zc3 add sysid set root_password=ZiitH.NOLOrRg set name_service=NONE set nfs4_domain=dynamic set security_policy=NONE set system_locale=C set terminal=xterms set timezone=US/Pacific end add node set physical-host=tm16-180 set hostname=rome add net set address=rome/24 set physical=nxge0 end end add node set physical-host=tm16-184 set hostname=bangalore add net set address=bangalore/24 set physical=e1000g0 end end add net set address=lh-3 end commit

15

MySQL conguration le

Sun Microsystems, Inc.

Appendix B

MySQL conguration le
MySQL master database conguration le: /pool2/data/my.cnf
[mysqld] server-id=1 #port=3306 #bind-address=pool2 socket=/tmp/lh-2.sock log=/pool2/data/logs/log1 log-bin=/pool2/data/logs/bin-log binlog-ignore-db=sc3_test_database log-slow-queries=/pool2/data/logs/log-slow-queries #log-update=/pool2/data/logs/log-update # Innodb #skip-innodb innodb_data_home_dir = /pool2/data/innodb innodb_data_file_path = ibdata1:10M:autoextend innodb_log_group_home_dir = /pool2/data/innodb #innodb_log_arch_dir = /pool2/data/innodb # You can set .._buffer_pool_size up to 50 - 80 % # of RAM but beware of setting memory usage too high set-variable = innodb_buffer_pool_size=50M set-variable = innodb_additional_mem_pool_size=20M # Set .._log_file_size to 25 % of buffer pool size set-variable = innodb_log_file_size=12M set-variable = innodb_log_buffer_size=4M innodb_flush_log_at_trx_commit=1 set-variable = innodb_lock_wait_timeout=50 # MySQL 4.x relay-log=/pool2/data/logs/slave-bin.log relay-log-info-file=/pool2/data/logs/slave-info

16

MySQL conguration le

Sun Microsystems, Inc.

MySQL slave database conguration le: /pool3/data/my.cnf

[mysqld] server-id=2 #port=3306 #bind-address=pool3 socket=/tmp/lh-3.sock log=/pool3/data/logs/log1 log-bin=/pool3/data/logs/bin-log binlog-ignore-db=sc3_test_database log-slow-queries=/pool3/data/logs/log-slow-queries #log-update=/pool3/data/logs/log-update # Innodb #skip-innodb innodb_data_home_dir = /pool3/data/innodb innodb_data_file_path = ibdata1:10M:autoextend innodb_log_group_home_dir = /pool3/data/innodb #innodb_log_arch_dir = /pool3/data/innodb # You can set .._buffer_pool_size up to 50 - 80 % # of RAM but beware of setting memory usage too high set-variable = innodb_buffer_pool_size=50M set-variable = innodb_additional_mem_pool_size=20M # Set .._log_file_size to 25 % of buffer pool size set-variable = innodb_log_file_size=12M set-variable = innodb_log_buffer_size=4M innodb_flush_log_at_trx_commit=1 set-variable = innodb_lock_wait_timeout=50 # MySQL 4.x relay-log=/pool3/data/logs/slave-bin.log relay-log-info-file=/pool3/data/logs/slave-info

17

Conguration les: /cong-les/mysql_cong

Sun Microsystems, Inc.

Appendix C

Conguration les: /cong-les/mysql_cong

From logical host lh-2 (master database):
# # # # # # Copyright 2003 Sun Microsystems, Inc. Use is subject to license terms. All rights reserved.

This file will be sourced in by mysql_register and the parameters listed below will be used.

# Where is mysql installed (BASEDIR) MYSQL_BASE=/usr/local/mysql # Mysql admin-user for localhost (Default is root) MYSQL_USER=root # Password for mysql admin user MYSQL_PASSWD=root # Configured logicalhost MYSQL_HOST=lh-2 # Specify a username for a faultmonitor user FMUSER=fmuser # Pick a password for that faultmonitor user FMPASS=fmuser # Socket name for mysqld ( Should be /tmp/<logical-host>.sock ) MYSQL_SOCK=/tmp/lh-2.sock # FOR SC3.1 ONLY, Specify the physical hostname for the # physical NIC that this logicalhostname belongs to for every node in the # cluster this Resourcegroup can located on. # IE: The logicalhost lh1 belongs to hme1 for physical-node phys-1 and # hme3 for physical-node phys-2. The hostname for hme1 is phys-1-hme0 and # for hme3 on phys-2 it is phys-2-hme3. # IE: MYSQL_NIC_HOSTNAME="phys-1-hme0 phys-2-hme3" MYSQL_NIC_HOSTNAME="paris mumbai" # where are your databases installed, (location of my.cnf) MYSQL_DATADIR=/pool2/data

18

Conguration les: /cong-les/mysql_cong

Sun Microsystems, Inc.

From logical host lh-3 (slave database):

# # # # # # Copyright 2003 Sun Microsystems, Inc. Use is subject to license terms. All rights reserved.

This file will be sourced in by mysql_register and the parameters listed below will be used.

# Where is mysql installed (BASEDIR) MYSQL_BASE=/usr/local/mysql # Mysql admin-user for localhost (Default is root) MYSQL_USER=root # Password for mysql admin user MYSQL_PASSWD=root # Configured logicalhost MYSQL_HOST=lh-3 # Specify a username for a faultmonitor user FMUSER=fmuser # Pick a password for that faultmonitor user FMPASS=fmuser # Socket name for mysqld ( Should be /tmp/<logical-host>.sock ) MYSQL_SOCK=/tmp/lh-3.sock # FOR SC3.1 ONLY, Specify the physical hostname for the # physical NIC that this logicalhostname belongs to for every node in the # cluster this Resourcegroup can located on. # IE: The logicalhost lh1 belongs to hme1 for physical-node phys-1 and # hme3 for physical-node phys-2. The hostname for hme1 is phys-1-hme0 and # for hme3 on phys-2 it is phys-2-hme3. # IE: MYSQL_NIC_HOSTNAME="phys-1-hme0 phys-2-hme3" MYSQL_NIC_HOSTNAME="rome bangalore" # where are your databases installed, (location of my.cnf) MYSQL_DATADIR=/pool3/data

19

Conguration les: /cong-les/ha_mysql_cong

Sun Microsystems, Inc.

Appendix D

Conguration les: /cong-les/ha_mysql_cong

From logical host lh-2 (master database):
# # # # # # # Copyright 2003 Sun Microsystems, Inc. Use is subject to license terms. All rights reserved.

This file will be sourced in by ha_mysql_register and the parameters listed below will be used.

# These parameters can be customized in (key=value) form # # RS - name of the resource for the application # RG - name of the resource group containing RS # PROJECT - A project in the zone, that will be used for this service # specify it if you have an su - in the start stop or probe, # or to define the smf credentials. If the variable is not set, # it will be translated as :default for the sm and default # for the zsh component # Optional # ZUSER - A user in the the zone which is used for the smf method # credentials. Yur smf servic e will run under this user # Optional # # BASEDIR # DATADIR # MYSQLUSER # LH # MYSQLHOST # FMUSER # FMPASS # LOGDIR # CHECK YES/NO. # HAS_RS # name of the Mysql bin directory name of the Mysql Data directory name of the user Mysql should be started of name of the LogicalHostname SC resource name of the host in /etc/hosts name of the Mysql fault monitor user name of the Mysql fault monitor user password name of the directory mysqld should store it's logfile. should HA-MySQL check MyISAM index files before start

- name of the MySQL HAStoragePlus SC resource

20

Conguration les: /cong-les/ha_mysql_cong

Sun Microsystems, Inc.

# # # # # # # # # # # # #

The following examples illustrate sample parameters for Mysql BASEDIR=/usr/local/mysql DATADIR=/global/mysqldata MYSQLUSER=mysql LH=mysqllh MYSQLHOST=mysqllh FMUSER=fmuser FMPASS=fmuser LOGDIR=/global/mysqldata/logs CHECK=YES

RS=master-mys-rs RG=mysql-1-rg PORT=22 LH=mysql-1-lh HAS_RS=mysql-1-hasp-rs # local zone specific options ZONE= ZONE_BT= ZUSER= PROJECT= # mysql specifications BASEDIR=/usr/local/mysql DATADIR=/pool2/data MYSQLUSER=mysql MYSQLHOST=lh-2 FMUSER=fmuser FMPASS=fmuser LOGDIR=/pool2/data/logs CHECK=YES

21

Conguration les: /cong-les/ha_mysql_cong

Sun Microsystems, Inc.

From logical host lh-3 (slave database):

# # # # # # # Copyright 2003 Sun Microsystems, Inc. Use is subject to license terms. All rights reserved.

This file will be sourced in by ha_mysql_register and the parameters listed below will be used.

# These parameters can be customized in (key=value) form # # RS - name of the resource for the application # RG - name of the resource group containing RS # PROJECT - A project in the zone, that will be used for this service # specify it if you have an su - in the start stop or probe, # or to define the smf credentials. If the variable is not set, # it will be translated as :default for the sm and default # for the zsh component # Optional # ZUSER - A user in the the zone which is used for the smf method # credentials. Yur smf servic e will run under this user # Optional # # BASEDIR # DATADIR # MYSQLUSER # LH # MYSQLHOST # FMUSER # FMPASS # LOGDIR # CHECK YES/NO. # HAS_RS # # # # # # # # # # # # # # name of the Mysql bin directory name of the Mysql Data directory name of the user Mysql should be started of name of the LogicalHostname SC resource name of the host in /etc/hosts name of the Mysql fault monitor user name of the Mysql fault monitor user password name of the directory mysqld should store it's logfile. should HA-MySQL check MyISAM index files before start

- name of the MySQL HAStoragePlus SC resource

The following examples illustrate sample parameters for Mysql BASEDIR=/usr/local/mysql DATADIR=/global/mysqldata MYSQLUSER=mysql LH=mysqllh MYSQLHOST=mysqllh FMUSER=fmuser FMPASS=fmuser LOGDIR=/global/mysqldata/logs CHECK=YES

22

Conguration les: /cong-les/ha_mysql_cong

Sun Microsystems, Inc.

RS=slave-mys-rs RG=mysql-2-rg PORT=22 LH=mysql-2-lh HAS_RS=mysql-2-hasp-rs # local zone specific options ZONE= ZONE_BT= ZUSER= PROJECT= # mysql specifications BASEDIR=/usr/local/mysql DATADIR=/pool3/data MYSQLUSER=mysql MYSQLHOST=lh-3 FMUSER=fmuser FMPASS=fmuser LOGDIR=/pool3/data/logs CHECK=YES

23

Conguration les: /cong-les/ha_mysql_cong

Sun Microsystems, Inc.

High Availability MySQL Database Replication with Solaris Zones Cluster

On the Web sun.com

Sun Microsystems, Inc. 4150 Network Circle, Santa Clara, CA 95054 USA Phone 1-650-960-1300 or 1-800-555-9SUN (9786) Web sun.com
2009 Sun Microsystems, Inc. All rights reserved. Sun, Sun Microsystems, the Sun logo, MySQL, StorEdge, and Sun BluePrints are trademarks or registered trademarks of Sun Microsystems, Inc. or its subsidiaries in the United States and other countries. All SPARC trademarks are used under license and are trademarks or registered trademarks of SPARC International, Inc. in the US and other countries. Products bearing SPARC trademarks are based upon an architecture developed by Sun Microsystems, Inc. Information subject to change without notice. Printed in USA 02/09