Upgrade to Oracle Solaris Cluster 3.3 When Using HA for MySQL

Use the information in this section to understand how to upgrade to Oracle Solaris Cluster 3.3 when using HA for MySQL.

Upgrade From Sun Cluster 3.1 8/05 to Oracle Solaris Cluster 3.3 When Using HA for MySQL

Use this procedure to upgrade HA for MySQL from Sun Cluster 3.1 8/05 to Oracle Solaris Cluster3.3.

1. Shut down the HA for MySQL resource.

   # clresource disableMySQL-resource

2. Upgrade the nodes to Oracle Solaris Cluster 3.3 according to Oracle Solaris Cluster Upgrade Guide.
3. Start the MySQL Server manually on Oracle Solaris Cluster 3.3.

   #cd MySQL-Base-directory
   
   # ./bin/mysqld --defaults-file=MySQL-Database-directory/my.cnf \
   --basedir=MySQL-Base-directory \
    --datadir=MySQL-Database-directory \
    --user=mysql \
    --pid-file=MySQL-Database-directory/mysqld.pid &

4. Access the MySQL instance from the local node with the socket option.

   # cd MySQL-Base-directory
   /bin/mysql -S MySQL-socket -uroot \
    -pAdminpassword \

   The following is an example for a MySQL instance.

   # mysql -s /tmp/hahostix1.sock -uroot -proot mysql> 

5. Drop the HA for MySQL test database sc3_test_database.

   # mysql -s /tmp/hahostix1.sock -uroot -proot
   mysql> DROP DATABASE sc3_test_database;
   Query OK, 0 rows affected (0.03 sec)

6. Delete all entries in db-table in mysql-database that contain User='MySQL-Faultmonitor-user'.

   # mysql -s /tmp/hahostix1.sock -uroot -proot mysql> use mysql; Database changed mysql> DELETE FROM db WHERE User='fmuser'; Query OK, 1 row affected (0.03 sec)

7. Delete all entries in user-table in mysql-database that contain User='<MySQL Faultmonitor user>'.

   # mysql -s /tmp/hahostix1.sock -uroot -proot mysql> use mysql; Database changed mysql> DELETE FROM user WHERE User='fmuser'; Query OK, 1 row affected (0.03 sec)

8. Add a fault-monitor user and a test database to MySQL.
   1. Navigate to the util directory.

      # cd /opt/SUNWscmys/util

   2. Copy the mysql_config to home-dir and edit the mysql_config file and follow the comments within that file.

      # Where is MySQL installed (BASEDIR)
      MYSQL_BASE=
      
      # MySQL admin-user for localhost (Should be root)
      MYSQL_USER=
      
      # Password for MySQL admin user
      MYSQL_PASSWD=
      
      # Configured logicalhost
      MYSQL_HOST=
      
      # Specify a username for a faultmonitor user
      FMUSER=
      
      # Pick a password for that faultmonitor user
      FMPASS=
      
      # Socket name for mysqld ( Should be /tmp/<Logical host>.sock )
      MYSQL_SOCK=
      
      # Specify the physical hostname for the physical NIC that this logicalhostname 
      # belongs to for every node in the cluster this resource group can be located on.
      # If you use the mysql_geocontrol features to implement the MySQL replication as 
      # the replication protocol in Oracle Solaris Cluster geographic edition, specify all
      # physical nodes of all clusters. Specify at least all the nodes on both sites
      # where the MySQL databases can be hosted.
      # 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"
      # IE: If two clusters are tied together by the mysql_geocontrol features, assuming the
      # mysql database on cluster one belongs to cl1-phys1-hme0 and cl1-phys2-hme3, the
      # mysql database on cluster two belongs to cl2-phys1-hme2 and cl2-phys2-hme4. Then the  
      # MYSQL_NIC_HOSTNAME variable needs to be set to:
      # MYSQL_NIC_HOSTNAME="cl1-phys1-hme0 cl1-phys2-hme3 cl2-phys1-hme2 cl2-phys2-hme4"
      
      MYSQL_NIC_HOSTNAME=
      
      # Where are your databases installed? (location of my.cnf)
      MYSQL_DATADIR=
      
      # Is MySQL Cluster installed?
      # Any entry here triggers the ndb engine check preparation. If no MySQL cluster should be
      # checked, leave it empty.
      NDB_CHECK=
      
      If you want to monitor the ndb tables of a MySQL cluster set NDB_CHECK to yes.

   The following is an example for a MySQL instance.

   MYSQL_BASE=/global/mysql
   MYSQL_USER=root
   MYSQL_PASSWD=root
   MYSQL_HOST=hahostix1
   FMUSER=fmuser
   FMPASS=fmuser
   MYSQL_SOCK=/tmp/hahostix1.sock
   MYSQL_NIC_HOSTNAME="clusterix1 clusterix2"
   MYSQL_DATADIR=/global/mysql-data
   NDB_CHECK=

9. After editing mysql_config, run the mysql_register script.

   # ./mysql_register -f home-dir/mysql_config

10. Stop the MySQL Server manually.

    # kill -TERM `cat MySQL-Database-directory/mysqld.pid

11. Start the HA for MySQL resource.

    # clresource enable MySQL-resource