git clone https://github.com/apache/shardingsphere.git;
cd shardingsphere;
mvn clean install -Prelease;
发布包所在目录为:/sharding-distribution/sharding-scaling-distribution/target/apache-shardingsphere-${latest.release.version}-sharding-scaling-bin.tar.gz
。
conf/server.yaml
,这里主要修改启动端口,保证不与本机其他端口冲突,其他值保持默认即可:port: 8888
blockQueueSize: 10000
pushTimeout: 1000
workerThread: 30
sh bin/start.sh
注意: 如果后端连接MySQL数据库,需要下载MySQL Connector/J, 解压缩后,将mysql-connector-java-5.1.47.jar拷贝到${sharding-scaling}\lib目录。
logs/stdout.log
,确保启动成功。Sharding-Scaling提供相应的HTTP接口来管理迁移任务,部署启动成功后,我们可以调用相应的接口来启动迁移任务。
创建迁移任务:
curl -X POST \
http://localhost:8888/shardingscaling/job/start \
-H 'content-type: application/json' \
-d '{
"ruleConfiguration": {
"sourceDatasource": "ds_0: !!org.apache.shardingsphere.orchestration.core.configuration.YamlDataSourceConfiguration\n dataSourceClassName: com.zaxxer.hikari.HikariDataSource\n properties:\n jdbcUrl: jdbc:mysql://127.0.0.1:3306/test?serverTimezone=UTC&useSSL=false\n username: root\n password: '\''123456'\''\n connectionTimeout: 30000\n idleTimeout: 60000\n maxLifetime: 1800000\n maxPoolSize: 50\n minPoolSize: 1\n maintenanceIntervalMilliseconds: 30000\n readOnly: false\n",
"sourceRule": "defaultDatabaseStrategy:\n inline:\n algorithmExpression: ds_${user_id % 2}\n shardingColumn: user_id\ntables:\n t1:\n actualDataNodes: ds_0.t1\n keyGenerator:\n column: order_id\n type: SNOWFLAKE\n logicTable: t1\n tableStrategy:\n inline:\n algorithmExpression: t1\n shardingColumn: order_id\n t2:\n actualDataNodes: ds_0.t2\n keyGenerator:\n column: order_item_id\n type: SNOWFLAKE\n logicTable: t2\n tableStrategy:\n inline:\n algorithmExpression: t2\n shardingColumn: order_id\n",
"destinationDataSources": {
"name": "dt_0",
"password": "123456",
"url": "jdbc:mysql://127.0.0.1:3306/test2?serverTimezone=UTC&useSSL=false",
"username": "root"
}
},
"jobConfiguration": {
"concurrency": 3
}
}'
注意:上述需要修改ruleConfiguration.sourceDatasource
和ruleConfiguration.sourceRule
,分别为源端ShardingSphere数据源和数据表规则相关配置;
以及ruleConfiguration.destinationDataSources
中目标端sharding-proxy的相关信息。
返回如下信息,表示任务创建成功:
{
"success": true,
"errorCode": 0,
"errorMsg": null,
"model": null
}
需要注意的是,目前Sharding-Scaling任务创建成功后,便会自动运行,进行数据的迁移。
执行如下命令获取当前所有迁移任务:
curl -X GET \
http://localhost:8888/shardingscaling/job/list
返回示例如下:
{
"success": true,
"errorCode": 0,
"model": [
{
"jobId": 1,
"jobName": "Local Sharding Scaling Job",
"status": "RUNNING"
}
]
}
进一步查询任务具体迁移状态:
curl -X GET \
http://localhost:8888/shardingscaling/job/progress/1
返回任务详细信息如下:
{
"success": true,
"errorCode": 0,
"errorMsg": null,
"model": {
"id": 1,
"jobName": "Local Sharding Scaling Job",
"status": "RUNNING"
"syncTaskProgress": [{
"id": "127.0.0.1-3306-test",
"status": "SYNCHRONIZE_REALTIME_DATA",
"historySyncTaskProgress": [{
"id": "history-test-t1#0",
"estimatedRows": 41147,
"syncedRows": 41147
}, {
"id": "history-test-t1#1",
"estimatedRows": 42917,
"syncedRows": 42917
}, {
"id": "history-test-t1#2",
"estimatedRows": 43543,
"syncedRows": 43543
}, {
"id": "history-test-t2#0",
"estimatedRows": 39679,
"syncedRows": 39679
}, {
"id": "history-test-t2#1",
"estimatedRows": 41483,
"syncedRows": 41483
}, {
"id": "history-test-t2#2",
"estimatedRows": 42107,
"syncedRows": 42107
}],
"realTimeSyncTaskProgress": {
"id": "realtime-test",
"delayMillisecond": 1576563771372,
"logPosition": {
"filename": "ON.000007",
"position": 177532875,
"serverId": 0
}
}
}]
}
}
数据迁移完成后,我们可以调用接口结束任务:
curl -X POST \
http://localhost:8888/shardingscaling/job/stop \
-H 'content-type: application/json' \
-d '{
"jobId":1
}'
返回如下信息表示任务成功结束:
{
"success": true,
"errorCode": 0,
"errorMsg": null,
"model": null
}
sh bin/stop.sh