Overview

This chapter describes CYCLONE, LOGMIRROR and CYFILE.

GOLDILOCKS supports two types of replication, and they are CYCLONE and LOGMIRROR. CYCLONE replicates transactions using CDC, and LOGMIRROR replicates redo log files from the source database.

CYFILE employs the CDC method to store or record transactions applied to the original database in Comma-Separated Values (CSV) format file.

• CYCLONE

  • It uses the Change Data Capture (CDC) method to analyze and process the redo log files of the source database, and then applies them to a remote database.

  • It only supports the asynchronous (async) method, as it analyzes the contents stored in the database's redo log file.

• LOGMIRROR

  • It replicates the redo log files stored in the source database to a remote database.

  • It is used to prevent data loss for CYCLONE, which operates in asynchronous mode.

• CYFILE

  • It analyzes the database's redo log file using the CDC method and stores the data in a CSV format file.

  • A user can perform replication using the corresponding file or carry out ETL based on the development's purpose.

Operating Method

CYCLONE

For more information about general operational methods and options, refer to CYCLONE.

Adding and Deleting Nodes

CYCLONE operates in group units, similar to the replication nodes. When adding nodes, a group must be added, and when deleting nodes, a group must be dropped.

Examples of Adding Nodes

• Record the group 2 to be added in the master configuration file.

  • Set a unique port for each group.

  • The default master configuration file is $GOLDILOCKS_DATA/conf/cyclone.master.conf.

• N/A

  • The following is an example of a group 1 node in operation.

...
...
GROUP_NAME = Group1
{
    PORT = 21102
    CAPTURE_TABLE =
    (
        testTable1,
        testTable2
    )
}

• N/A

  • The following is an example of a group 2 node to be added.

GROUP_NAME = Group2
{
    PORT = 21103
    CAPTURE_TABLE =
    (
        testTable5,
        testTable6
    )
}

• Record the group 2 to be added in the slave configuration file.

  • The port for group 2 must be the same as the port added to the existing master.

  • The default slave configuration file is $GOLDILOCKS_DATA/conf/cyclone.slave.conf.

• N/A

  • The following is an example of a group 1 node in operation.

...
...
GROUP_NAME = Group1
{
    PORT = 21102    
    APPLY_TABLE = 
    (
        testTable1 To testTable3,
        testTable2 To testTable4
    )
}

• N/A

  • The following is an example of a group 2 node to be added.

GROUP_NAME = Group2
{
    PORT = 21103    
    APPLY_TABLE = 
    (
        testTable5 To testTable7,
        testTable6 To testTable8
    )
}

• Execute and verify the added group 2 node on the master device.

prompt> cyclone --master --start --group Group2
[GROUP2] Startup done as Master.

prompt> cyclone --master --status
======================================
|       CYCLONE STATUS - MASTER      |
======================================
 GROUP1 Running...
 GROUP2 Running...
--------------------------------------

• Execute and verify the added group 2 node on the slave device.

prompt> cyclone --slave --start --group Group2
[GROUP2] Startup done as Slave.

prompt> cyclone --slave --status
======================================
        CYCLONE STATUS - SLAVE        
======================================
 GROUP1 Running...
 GROUP2 Running...
--------------------------------------

Examples of Deleting Nodes

• Terminate the group 2 node to be deleted on the slave device as follows.

prompt> cyclone --slave --stop --group Group2
stop done.

prompt> cyclone --slave --status
======================================
CYCLONE STATUS - SLAVE
======================================
GROUP1 Running...
--------------------------------------

• Terminate the group 2 node to be deleted on the master device as follows.

prompt> cyclone --master --stop --group Group2
stop done.

prompt> cyclone --master --status
======================================
CYCLONE STATUS - MASTER
======================================
GROUP1 Running...
--------------------------------------

• Drop the group 2 node to be deleted from the master configuration file.

  • The default master configuration file is $GOLDILOCKS_DATA/conf/cyclone.master.conf.

• N/A

  • The following is an example of a group 1 node in operation.

...
...
GROUP_NAME = Group1
{
    PORT = 21102
    CAPTURE_TABLE =
    (
        testTable1,
        testTable2
    )
}

• N/A

  • The following is an example of how to delete the group 2 node.

GROUP_NAME = Group2
{
    PORT = 21103
    CAPTURE_TABLE =
    (
        testTable5,
        testTable6
    )
}

• Drop the group 2 node to be deleted from the slave configuration file as follows.

  • The default slave configuration file is $GOLDILOCKS_DATA/conf/cyclone.slave.conf.

• N/A

  • The following is an example of a group 1 node in operation.

...
...
GROUP_NAME = Group1
{
    PORT = 21102    
    APPLY_TABLE = 
    (
        testTable1 To testTable3,
        testTable2 To testTable4
    )
}

• N/A

  • The following is an example of how to delete the group 2 node.

GROUP_NAME = Group2
{
    PORT = 21103    
    APPLY_TABLE = 
    (
        testTable5 To testTable7,
        testTable6 To testTable8
    )
}

Initializing Replication

Initializing replication is executed when existing replication nodes or a group's table gives up execution due to a DDL operation. Either a specific node or all nodes can be initialized.

The initializing replication is performed by restarting the replication running on the slave using the --reset option.
On the other hand, no action is required on the master.

Examples of Initializing Replication on a Specific Node

• Terminate the group 2 node to be initialized on the slave device as follows.

prompt> cyclone --slave --stop --group Group2
stop done.

prompt> cyclone --slave --status
======================================
        CYCLONE STATUS - SLAVE        
======================================
 GROUP1 Running...
--------------------------------------

• Restart the group 2 node on the slave device using the --reset option.

  • No action is required on the master device.

  • Replication will restart from the current point.

prompt> cyclone --slave --start --reset --group Group2
[GROUP2] Startup done as Slave.

prompt> cyclone --slave --status
======================================
        CYCLONE STATUS - SLAVE        
======================================
 GROUP1 Running...
 GROUP2 Running...
--------------------------------------

Examples of Initializing Replication for All Nodes

• Terminate all CYCLONE operations running on the slave device.

prompt> cyclone --stop --slave

• Restart CYCLONE on the slave device using the --reset option.

  • No action is required on the master device.

  • Replication will resume from the current point.

prompt> cyclone --slave --start --reset
[GROUP1] Startup done as Slave.
[GROUP2] Startup done as Slave.

prompt> cyclone --slave --status
======================================
        CYCLONE STATUS - SLAVE        
======================================
 GROUP1 Running...
 GROUP2 Running...
--------------------------------------

LOGMIRROR

For more information about general operational methods and options, refer to LOGMIRROR.

Retrieving LOGMIRROR State

When interworking with LOGMIRROR, GOLDILOCKS includes the response waiting process of LOGMIRROR. If LOGMIRROR is waiting for a response, GOLDILOCKS will also be in a blocked state, waiting. This state can be checked in v$system_stat.

gSQL> SELECT * FROM V$SYSTEM_STAT WHERE STAT_NAME='LOG_MIRROR_SYNC_STATE';

STAT_NAME             STAT_VALUE COMMENTS                                     
--------------------- ------ ---------------------------------------------
LOG_MIRROR_SYNC_STATE      0 logmirror sync state( 0 : sync, 1 : blocked )


1 row selected.

If STAT_VALUE is 0, it indicates an ordinary state rather than a standby state. If STAT_VALUE is 1, it indicates a blocked state waiting for a response. To restart the GOLDILOCKS service while LOGMIRROR is waiting for a response, the LOGMIRROR service can be stopped by modifying LOG_MIRROR_TIMEOUT as follows.

gSQL> ALTER SYSTEM SET LOG_MIRROR_TIMEOUT = 20;

System altered.

Initializing Replication

Initializing the replication of LOGMIRROR must be done manually to prevent data loss or an unrecoverable situation caused by incorrect user options.

Examples of Initializing Replication

• Terminate LOGMIRROR in operation on the slave device as follows.

prompt> logmirror --slave --stop 
stop done.

• Terminate LOGMIRROR in operation on the master device as follows.

prompt> logmirror --master --stop 
stop done.

• Drop the replicated control files and redo log files from the slave device.

  • For more information about the path, refer to the 'LOG_PATH' option described in the slave configuration file.

  • The default LOGMIRROR slave configuration file: $GOLDILOCKS_DATA/conf/logmirror.slave.conf.

CYFILE

For more information about general operational methods and options, refer to CYFILE.

Starting, Stopping, and Checking the Status of Cyfile

Example

• Start cyfile from the current point using the default configuration file.

prompt> cyfile --start --reset all
Startup done.

• Stop cyfile.

prompt> cyfile --stop
Stop done.

• CYFILE operates in groups, and the data files are created in groups.

prompt> cyfile --status
cyfile --status
======================================
|          CYFILE STATUS             |
======================================
 GROUP1 Running...
======================================

Trace Log

The following provides detailed information about the trace log.

Troubleshooting for CYCLONE

The following are error messages and troubleshooting for CYCLONE.

Troubleshooting for LOGMIRROR

The following are error messages and troubleshooting for LOGMIRROR.

Troubleshooting for CYFILE

The following are error messages and troubleshooting for CYFILE.