Tungsten replicator do mysql cluster backup

About Tungsten replicator

Tungsten is a family of open source technologies for database clustering and replication. Tungsten includes replication, management, SQL routing, and proxying that improve database availability, protect data, and raise application throughput.

Download:http://sourceforge.net/projects/tungsten/files/

tungsten replicator is the software for data replication in a component. Adopted in the main from the server to deploy the java process of the binlog on master database to track and update the contents of resolution sent to them from the server, and the java from the server will receive a written update from the database.

Test
Master1
192.168.1.8
Master2
192.168.1.4
Slave
192.168.1.7

Master configuration

 
Tungsten replicator Download compressed package, unzip to a separate directory. Conf directory will be renamed to replicator.properties.mysql the replicator.properties, and in accordance with the following practical situation:

# Master server is the statement of the current model
replicator.role = master
# Services start automatically after the copy
replicator.auto_enable = true
# Specify a unique global id, server name proposed
replicator.source_id = dev08
# Copy the data designated for storage of the database name manually in the mysql database to generate a space and this library will my.cnf by adding binlog-ignore-db parameter in. Default database for the tungsten
replicator.schema = tungsten
# Designated master of ip, this machine is the use of localhost
replicator.thl.remote_uri = thl: //localhost/
# If replicator.schema parameter specifies the database to other name, would be taken into tungsten database name specified.
replicator.thl.url = jdbc: mysql://localhost/tungsten
# Specify the binlog file directory
replicator.extractor.mysql.binlog_dir =/usr/local/mysql/var
# Specify the binlog file name prefix
replicator.extractor.mysql.binlog_file_pattern = mysql-bin
# Replicator procedures specified database user name and password, and in the database to generate the user, the need for super powers.
replicator.thl.user=tungsten
replicator.thl.password=secret
replicator.applier.mysql.host=localhost
replicator.applier.mysql.port=3306
replicator.applier.mysql.user=tungsten
replicator.applier.mysql.password=secret
replicator.extractor.mysql.host=localhost
replicator.extractor.mysql.user=tungsten
replicator.extractor.mysql.password=secret

Upon completion of the use of configuration directory under the bin command to start trepsvc
Trepsvc start
Bin directory with view server state trepctl
Trepctl status
Name
Value

===============================================
System ID
dev08
System Version:
1.0-beta5
System State:
ONLINE:MASTER
System Uptime (S): 282739.972s
State Uptime (S):
282739.351s
Error:
null
Error Exception:
null
Min Seq No:
0
Max Seq No:
179
Monitor Intvl (S): 282740.011
Extr Total:
179
Extr Last Seq No:
179
Extr/Sec:
6.330904427029223E-4
Recv Total:
0
Recv Last Seq No:
-1
Recv Source TS:
null
Recv Target TS:
null
Recv Latency (S):
0.0
Recv/Sec:
0.0
Apply Total:
0
Apply Last Seq No: -1
Apply Source TS:
null
Apply Target TS:
null
Apply Latency (S): 0.0
Apply/Sec:
0.0
State: ONLINE:MASTER
Seqno Range: 0 -> 179
When the state is ONLINE: MASTER note start successfully

Slave configuration

From tow master as a result of the synchronous data, so to extract tungste replicator into two different directory, then start the two directory tungsten replicator process.
And master the basic configuration is the same as the contents of replicator.properties is different from the actual situation in accordance with amendments to the following:

# Statement slave mode the machine is
replicator.role = slave
# Services start automatically after the copy
replicator.auto_enable = true
# Specify a unique global id, slave on two different replicator need id
replicator.source_id = dev071
# Copy the data designated for storage of the library and the library will be in my.cnf by adding binlog-ignore-db parameter in. The need for two different replicator release inventory data, generated in the mysql manual in which two empty Treasury, and in the my.cnf to join two database binlog-ignore-db parameter in
replicator.schema = svnrep
# Rmi ports, two need to specify a different port replicator
replicator.rmi_port = 11000
# Thl ports, two need to specify a different port replicator
replicator.thl.uri = thl:// 0.0.0.0:12112/
# Appointed master of the ip, two replicator need to specify a different master for the replication database.
replicator.thl.remote_uri = thl://192.168.1.4/
# Replicator.schema parameters, will be replaced by the actual svnrep database name.
replicator.thl.url = jdbc: mysql://localhost/svnrep
Other configurations with the same master.

Upon completion of the use of configuration directory under the bin command to start trepsvc
Trepsvc start
Bin directory with view server state trepctl
Trepctl status
Name
Value

===============================================
System ID
dev071
System Version:
1.0-beta5
System State:
ONLINE:SLAVE
System Uptime (S): 6509.149s
State Uptime (S):
6325.891s
Error:
null
Error Exception:
null
Min Seq No:
0
Max Seq No:
5
Monitor Intvl (S): 6509.153
Extr Total:
0
Extr Last Seq No:
-1
Extr/Sec:
0.0
Recv Total:
2
Recv Last Seq No:
5
Recv Source TS:
2009-06-22 15:53:54.581
Recv Target TS:
2009-06-22 15:58:26.343
Recv Latency (S):
271.762
Recv/Sec:
3.072588835840643E-4
Apply Total:
1
Apply Last Seq No: 5
Apply Source TS:
2009-06-22 15:53:54.581
Apply Target TS:
2009-06-22 15:58:26.352
Apply Latency (S): 271.771
Apply/Sec:
1.536292765781721E-4
State: ONLINE:SLAVE
Seqno Range: 0 -> 5
When the state is ONLINE: SLAVE note start successfully