1.
Application shutdown
2. Run on the master server: [Link]:
mysql> flush tables with read lock;
mysql> set global read_only=1; set global super_read_only=1;
mysql> show master status;
3. Note the file name and position:
master@[Link]: (none)>show master status;
+------------------+-----------+--------------+------------------
+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
Executed_Gtid_Set |
+------------------+-----------+--------------+------------------
+-------------------+
| mysql-bin.338622 | 262776281 | | |
|
+------------------+-----------+--------------+------------------
+-------------------+
1 row in set (0.00 sec)
master@[Link]: (none)>
4. Allow all the replicas to sync to the same file and position. Run the following
on each of the following hosts to check sync status:
[Link]
[Link]
[Link]
mysql> show slave status;
Relay_Master_Log_File: mysql-bin.338258
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Replicate_Do_DB:
Replicate_Ignore_DB:
Replicate_Do_Table:
Replicate_Ignore_Table:
Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 876089268
5. Once the replicas have all synced up, run the following commands on each of them
to perform a clean shutdown:
mysql> set global innodb_fast_shutdown=0;
$ sudo systemctl stop mysqld
6. Run the following commands on the master ([Link]):
mysql> set global innodb_fast_shutdown=0;
$ sudo systemctl stop mysqld
7. Adjust /etc/[Link] of the master server ([Link]) to with the
configuration below:
expire_logs_days=0
read_only=1
super_read_only=1
8. Poweroff the following set of hosts:
[Link]
[Link]
[Link]
9. Rename the hosts as follows:
svdchpemakerdb01 ([Link]) -> svdcpmakerdb02 ([Link])
svdchpemakerdb02 ([Link]) -> svdcpmakerdb03 ([Link])
svdrpmakerdb01 ([Link]) -> svdrpmakerdbstby ([Link])
10. Create new /etc/[Link] files on each of the hosts above. Run the following set
of commands on the hosts [Link], [Link], [Link]
$ mv /etc/[Link] /etc/my.cnf_08042025_before_switchover
$ cp -a /app/bpmsdatabase/mysql/[Link] /etc/[Link]
10. Restart mysql in READONLY mode on the following hosts: [Link],
[Link], [Link]
$ sudo systemctl start mysqld
11. On the master node ([Link]), run the command below and note the file
name and position:
mysql> show master status;
master@[Link]: (none)>show master status;
+---------------------------+----------+--------------+------------------
+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
Executed_Gtid_Set |
+---------------------------+----------+--------------+------------------
+-------------------+
| svdcpmakerdb02-bin.000001 | 157 | | |
|
+---------------------------+----------+--------------+------------------
+-------------------+
12. On each of the replicas ([Link], [Link]), run the following
commands:
mysql> reset slave;
mysql>
CHANGE MASTER TO MASTER_HOST='[Link]', MASTER_USER='repl',
MASTER_PASSWORD='R3pl1c9t1_tbd0n', MASTER_LOG_FILE='svdcpmakerdb02-bin.000001',
MASTER_LOG_POS=157;
mysql> start slave;
13. Check the sync between the master and the replica by running the following on
hosts [Link] and [Link]:
mysql> show slave status;
Relay_Master_Log_File: mysql-bin.338258
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Replicate_Do_DB:
Replicate_Ignore_DB:
Replicate_Do_Table:
Replicate_Ignore_Table:
Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 876089268
14. If the syncing is okay, put the master host ([Link]) in READ-WRITE
mode:
mysql> set global read_only=0; set global super_read_only=0;
15. Comment the read_only and super_read_only variables in the /etc/[Link] of the
master host [Link] by prepending the variables with a hash symbol
16. Start the application and test
17. Comment the skip_slave_start variable in the /etc/[Link] of the replicas
([Link] and [Link])
19. Map the ZFS mountpoint [Link]:/export/BFUBPRD/pmakerapp to /backup on
the following hosts [Link], [Link], and [Link]
20. Monitor the database and application for any anomalies
