java程序跨服务器跨数据库批量导入导出百万级数据

java程序跨服务器跨数据库批量导入导出百万级数据

/**@author G007N
  * 日期：2012年4月4日
  * java程序跨服务器跨数据库批量导入导出百万级数据
  * 本例主要是从sqlserver 2008数据库导入数据到oracle 10.2.0数据库
  */
 
 public static void main(String args[]) throws Exception {
  ResultSet rs = null;
  Statement s1 = null;
  Statement s2 = null;
  Connection cn1 = null;
  Connection cn2 = null;
  long startTime = 0;//开始时间
  long endTime = 0;  //结束时间
  int count = 1;     //计数
  int onerun = 0;    //执行的最大数
  int datanum = 0;   //总条数
  int num = 0;       //取整最大数据
  
  try {
   startTime = System.currentTimeMillis();
   Class.forName("com.microsoft.sqlserver.jdbc.SQLServerDriver");
   Class.forName("oracle.jdbc.driver.OracleDriver");
   cn1 = DriverManager.getConnection("jdbc:oracle:thin:@192.168.1.16:1521:oracleDb","test","test2012");
   cn2 = DriverManager.getConnection("jdbc:sqlserver://192.168.1.18:1433;databaseName=dbsql;user=sa;password=db2012");
   cn1.setAutoCommit(false);//插入oracle数据库时使用事务批量提交
   if (cn2 != null) {
    s1 = cn1.createStatement();
    s2 = cn2.createStatement(ResultSet.TYPE_SCROLL_SENSITIVE,ResultSet.CONCUR_UPDATABLE);
    rs = s2.executeQuery("select count(vcBusiFlowNo) from tb_His2_Ask_Pay_Fee");
    if (rs.next()) {//先查询sqlserver数据源表，获取总数据条数
     datanum = rs.getInt(1);
    }
    System.out.println("总记录数："+datanum+"条");
    onerun = 10000;//执行的最大数
    num = datanum/onerun*onerun;//计算出总条数符合每批10000的数量是多少
    s2.setMaxRows(datanum);
    s2.setFetchSize(onerun);//每批执行的数据条数
    rs = s2.executeQuery("select nSerialNo,vcBusiFlowNo,vcBusiTypeID,vcBusiNumber,vcUserName,vcUserAddress,vcUserLinkTel,nPayFee,vcBusiMonth,vcComments,dOperDate,vcOperID,vcDealerID,vcAreaCode,vcDealerGradeID from tb_His2_Ask_Pay_Fee");
    while (rs.next()) {
     s1.addBatch("insert into tb_His2_Ask_Pay_Fee(nSerialNo,vcBusiFlowNo,vcBusiTypeID,vcBusiNumber,vcUserName,vcUserAddress,vcUserLinkTel,nPayFee,vcBusiMonth,vcComments,dOperDate,vcOperID,vcDealerID,vcAreaCode,vcDealerGradeID) values('"
            +rs.getString("nSerialNo")+"','"
            +rs.getString("vcBusiFlowNo")+"','"
            +rs.getString("vcBusiTypeID")+"','"
            +rs.getString("vcBusiNumber")+"','"
            +rs.getString("vcUserName")+"','"
            +rs.getString("vcUserAddress")+"','"
            +rs.getString("vcUserLinkTel")+"','"
            +rs.getString("nPayFee")+"','"
            +rs.getString("vcBusiMonth")+"','"
            +rs.getString("vcComments")+"','"
            +rs.getString("dOperDate")+"','"
            +rs.getString("vcOperID")+"','"
            +rs.getString("vcDealerID")+"','"
            +rs.getString("vcAreaCode")+"','"
            +rs.getString("vcDealerGradeID")+"')");
     if (count > num) {//10000取整后剩余的小数据量就顺序插入
      s1.executeBatch();
      cn1.commit();
      s1.clearBatch();
     } else {//数据够批次的就按批量插入
      if (count % onerun == 0) {//10000条一批插入
       s1.executeBatch();
       cn1.commit();
       s1.clearBatch();
      }
     }
     count++;
    }
   }
  } catch (Exception e) {
   e.printStackTrace();
  } finally {
   rs.close();
   s2.close();
   s1.close();
   cn2.close();
   cn1.close();
  }
  endTime = System.currentTimeMillis();
  System.out.println("成功移植数据："+(count-1)+"条，耗时"+(endTime-startTime)/1000+"秒");
 }
 
执行结果如下： 
总记录数：2268422条
成功移植数据：2268422条，耗时6346秒