我有一个类似下面的查询,并想知道通过批处理PreparedStatement生成什么样的SQL。
INSERT INTO table1 (id, version, data)
VALUES (?, ?, ?)
ON DUPLICATE KEY UPDATE
table1.data = IF(table1.version > table2.version, table1.data, table2.data),
table1.version = IF(table1.version > table2.version, table1.version, table2.version)
问题是,它是否会将此解析为批处理中每行的整个sql字符串的副本,或者它是否会执行以下操作:
INSERT INTO table1 (id, version, data)
VALUES (a1, b1, c1), (a2, b2, c2), (a3, b3, c3), ...
ON DUPLICATE KEY UPDATE
table1.data = IF(table1.version > table2.version, table1.data, table2.data),
table1.version = IF(table1.version > table2.version, table1.version, table2.version)
如果没有,那么性能含义是什么?如何编写它以便我可以使用PreparedStatement批处理这些INSERT..UPDATE语句而不会导致性能损失?
答案 0 :(得分:2)
准备好的语句只是将您放入的位置值插入到重复语句中,然后每次都不需要解析。因此,您的第二种形式只需要N * 3参数,并且不会为您提供准备好的语句的任何速度提升。对于重复语句,您要使用addTobatch。基本上你准备语句,(例如“更新......??”然后一次添加3个参数,并一次执行批处理。
我曾经使用类似这样的东西来包装它的混乱。所以你只需做类似
的事情 SQLBatchHandler h = new SQLBatchHandler(conn, "UPDATE ... WHERE ? ? ? ... ");
h.addToBatch(x, y,z);
h.addToBatch(x2,y2,z2);
...
h.flush();
public class SQLBatchHandler {
public static int MAX_BATCH_SIZE = 500;
public String query;
private Connection conn;
private PreparedStatement ps;
private int batch_ct;
public SQLBatchHandler(Connection c, String query) throws SQLException
{
conn = c;
this.query = query;
ps = conn.prepareStatement(query);
}
/**
* add this row to the batch and handle the commit if the batch size
* exceeds {@link #MAX_BATCH_SIZE}
*
* @param values row values
* @throws SQLException
*/
public void addToBatch(Object ... values) throws SQLException
{
int i = 0;
for (Object value: values)
{
ps.setObject((++i), value);
}
add();
}
private void add() throws SQLException
{
ps.addBatch();
if ((++batch_ct) > MAX_BATCH_SIZE)
{
ps.executeBatch();
batch_ct = 0;
}
}
/**
* Commit any remaining objects and close.
*
* @throws SQLException On statement close error.
*/
public void flush() throws SQLException
{
if (batch_ct == 0) { return; }
try
{
ps.executeBatch();
}
catch (SQLException e)
{
throw e;
}
finally
{
if (ps != null)
{
ps.close();
}
}
}
}