Too many busy connections using JDBCTemplate and c3p0

I am developing a web application with database access using Spring, JDBCTemplate and c3p0.

I often have a server freeze, and I am pretty sure it comes from the number of busy database connections. If I watch the application behavior, using jconsole, I can see that the maxPoolSize of the ComboPooledDataSource is reached, and the server doesn't load a page anymore.

Here is the useful code:

DataSource definition:

<Resource auth="Container" description="GDLWeb DB Connection"
    driverClass="org.postgresql.Driver"
    maxPoolSize="16"
    minPoolSize="1"
    acquireIncrement="1"
    maxIdleTime="60"
    maxStatements="0"
    idleConnectionTestPeriod="1800"
    acquireRetryAttempts="30"
    breakAfterAcquireFailure="true"
    name="jdbc/gdlweb"
    user="gdlweb"
    password=""
    factory="org.apache.naming.factory.BeanFactory"
    type="com.mchange.v2.c3p0.ComboPooledDataSource"
    jdbcUrl="jdbc:postgresql://localhost:5432/postgres"
/>

Typical access method (in DAO class):

protected T getPersistentObject(
        final String tableName,
        final List<WhereClause> whereParams,
        final RowMapper<T> rowMapper) {

    try {
        log.debug(this, "get " + tableName + " " + whereParams);
        return (T) getTemplate().queryForObject(
                generateSelectStar(tableName, whereParams),
                extractValueMap(whereParams),
                rowMapper);
    } catch (final EmptyResultDataAccessException e) {
        log.warning(this, "No " + tableName + " found with " + whereParams + " in the DB!");
        return null;
    }
}

I tried to increase the maxPoolSize to 100, which is the maxConnections defined in my postgresql server. This way, I could see that there were 43 busy connections currently openned, just before the postgresql server crashes.

I am probably using JDBCTem开发者_Go百科plate the wrong way, but I don't know where.

Thanks.

---

The problem may be with the Mysql Connector/J version you are using.

I had the same issue, updating to the new Mysql Connector v5.1.15 solved it for me. v5.1.13 has a bug which results in the problems you are seeing

Change-log for the version which fixes the bug: http://dev.mysql.com/doc/refman/5.1/en/cj-news-5-1-14.html

Thanks