jdbc 大数据批量插入很慢问题解决

阅读更多

 

 普通的执行过程是:每处理一条数据,就访问一次数据库;

而批处理是:累积到一定数量,再一次性提交到数据库,减少了与数据库的交互次数,所以效率会大大提高

 但是使用了 addBatch() -> executeBatch() 还是很慢

 

解决:

a.需配置 参数 在数据库连接URL后面加上这个参数:      

    dbUrl "jdbc:mysql://localhost:3306/User?rewriteBatchedStatements=true";

b.conn.setAutoCommit(false);    需要设置 手动提交

 

完整例子:

1.正常JDBC 版本

String sql = "insert into table *****";
con.setAutoCommit(false);
ps = con.prepareStatement(sql);
for(int i=1; i<65536; i++){
    ps.addBatch();
    // 1w条记录插入一次
    if (i % 10000 == 0){
         ps.executeBatch();
         con.commit();
     }
}
// 最后插入不足1w条的数据
ps.executeBatch();
con.commit();

 

 

2.Spring boot jdbc版本

 

  @Autowired

    private JdbcTemplate jdbcTemplate;

 

private void batchInsertOrUpdateGoodsTemp(List list){
        StringBuilder sql = new StringBuilder("INSERT INTO  `gds_goods_temp` (`CAT_CODE`, `BRAND_CODE`, `GOODS_SN`, `GOODS_STATUS`, `GOODS_NAME` ").append(
          ", `IS_MARKET`, `IS_DELETE`, `GOODS_INTRO`, `GOODS_DETAIL`, `UNIT`, `WEIGHT`, `SPEC`, `BARCODE`, `GOODS_IMG`, `GOODS_THUMB`, `GOODS_GALLERY_IMGS`")
          .append(", `MARKET_PRICE`, `REMARK`, `ADD_TYPE`, `ADD_MCH_NO`) VALUES(?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)");
        
        KeyHolder key = new GeneratedKeyHolder();
        this.jdbcTemplate.update(con -> {
            con.setAutoCommit(false);
            PreparedStatement preState =null;
            try {
                preState = con.prepareStatement(sql.toString(), Statement.RETURN_GENERATED_KEYS);
                for(GoodsTemp temp : list){
                    preState.setString(1, temp.getCatCode());
                    preState.setString(2, temp.getBrandCode());
                    preState.setString(3, temp.getGoodsSn());
                    preState.setString(4, temp.getGoodsStatus());
                    preState.setString(5, temp.getGoodsName());
                    preState.setString(6, temp.getIsMarket());
                    preState.setString(7, temp.getIsDelete());
                    preState.setString(8, temp.getGoodsIntro());
                    preState.setString(9, temp.getGoodsDetail());
                    preState.setString(10, temp.getUnit());
                    preState.setBigDecimal(11, temp.getWeight());
                    preState.setString(12, temp.getSpec());
                    preState.setString(13, temp.getBarcode());
                    preState.setString(14, temp.getGoodsImg());
                    preState.setString(15, temp.getGoodsThumb());
                    preState.setString(16, temp.getGoodsGalleryImgs());
                    preState.setBigDecimal(17, temp.getMarketPrice());
                    preState.setString(18, temp.getRemark());
                    preState.setString(19, temp.getAddType());
                    preState.setString(20, temp.getAddMchNo());
                    preState.addBatch();
                }
                preState.executeBatch();
                con.commit();
            } catch (SQLException e) {
                log.error("sql语句{},批量执行错误:{}", sql,e.getMessage());
            }
            return preState;
        }, key);
    }

 

 

 

 

 

你可能感兴趣的:(jdbc 大数据批量插入很慢问题解决)