Hikari: 2.4.7
PostgreSQL JDBC driver: 9.4-1201-jdbc41
I'm trying to understand what must be done to a java.sql.Connection object for it to be available again in the
connection pool?
I've just introduced connection pooling to a multi threaded application that was
previously standing up / tearing down connections with each SQL statement.
What I have noticed, after introducing Hikari, is that as soon as I hit maximumPoolSize every attempt
thereafter to HikariDataSource.getConnection will fail due to connectionTimeout. So it seems like I'm not "releasing" this connection somehow.
The typical use of the Connection object is:
# omits Exception handling, parameter substitution, result evaluation.
PreparedStatement preparedStatement = hikariDataSource.getConnection().prepareStatement(sql);
preparedStatement.executeQuery();
preparedStatement.close();
Is there anything else that is expected to be done on this connection to get it eligible for reuse in the connection pool?
Autocommit is on. Connection.close(), unless doing something special when provided by Hikari, seemed like the exact thing I wanted to avoid.
I don't know Hikari specifically, but for every connection you take out of a connection pool, you have to return that connection when you are done with it.
Typically this is done using Connection.close() - the pool hands out a wrapper function where close() doesn't physically close the connection, only returns it.
So your code should look like this:
Connection con = hikariDataSource.getConnection();
PreparedStatement preparedStatement = con.prepareStatement(sql);
preparedStatement.executeQuery();
preparedStatement.close();
con.close(); // this returns the connection to the pool
Of course the two close() methods should be called in a finally block.
Related
I'm getting random " The connection is closed: The connection is closed" errors when using Oracle UCP, v 12.1.0.2.0. It looks like connection is marked as closed in
oracle.ucp.jdbc.proxy.JDBCConnectionProxyFactory#invoke :
if(Clock.isBefore(this.creationTS, this.m_jdbcPooledConnection.getAvailableStartTime()) || Clock.isBefore(this.creationTS, this.m_jdbcPooledConnection.getBorrowedStartTime())) {
this.m_closed = Boolean.valueOf(true);
}
The Clock.isBefore(this.creationTS, this.m_jdbcPooledConnection.getAvailableStartTime()) returns true.
Could somebody please explain what this check is for?
The getAvailableStartTime is set when connection is retured to the pool, the creationTS - is set when JDBCConnectionProxyFactory is being created and it's being created when giving connection away.
The isBefore looks like this:
public static boolean isBefore(long time1, long time2) {
return time1 < time2 - 1000L;
}
So, is the condition for the cases when connection was returned less than a second ago?
ps: tried validation query "select 1 from dual" - no effect
If Clock.isBefore(this.creationTS, this.m_jdbcPooledConnection.getAvailableStartTime()) returns true then it means that UCP has recollected the connection and made it available again. This typically happens if you turn on connection harvesting in UCP. UCP detects when a connection is borrowed but not used for too long (poorly designed application) and to avoid connection leaks it will grab the connection back and make it available in the pool. If the original thread then wakes up and attempts to use the connection it gets a connection is closed error.
Got the db connection (conn) from the pool.
Assume that autocommit is TRUE on that connection.
Now conn.setautocommit(false) has set ;
then after few statement updates and finally conn.commit()/conn.rollback() has done.
Now do i need to do explicitly code setautocommit(true) to revert to the previous conn state?
OR commit()\rollback() will set setautocommit(true) inherently ?
That depends on where you got that connection from. If you created the connection yourself, there is no need to restore the state of auto commit.
If you got it from a data source, you should restore the state to what it was because the data source might keep the connections in a pool and the next piece of code might not expect what you set.
commit() doesn't influence the value of auto commit. Enabling auto commit just makes sure that the JDBC driver calls commit() after each statement that you execute. You can still call commit() as often as you like, it just won't have any effect (except that rollback() will not always do what you want).
[EDIT] How auto commit is handled depends on your connection pool. dbcp has a config option to turn auto commit off before giving you a connection, c3p0 will roll back connections when you return then to the pool. Read the documentation for your connection pool how it works.
If you don't know which pool is used, the safe solution is to set auto commit to false whenever you get a connection and to roll back the connection if you get an exception. I suggest to write a wrapper:
public <T> T withTransaction( TxCallback<T> closure ) throws Exception {
Connection conn = getConnection();
try {
boolean autoCommit = conn.getAutoCommit();
conn.setAutoCommit(false);
T result = closure.call(conn); // Business code
conn.commit();
conn.setAutoCommit(autoCommit);
} catch( Exception e ) {
conn.rollback();
} finally {
conn.close();
}
}
This code will correctly handle the connection for you and you don't need to worry about it anymore in your business code.
Interestingly, conn.setAutoCommit(true); implies a commit (if it's in autoCommit(false) mode, see here, but it might be clearer to people if you still break them out.
In Oracle 12c connection will be defaulted to the autocommit true. But if you set the autocommit as false, you need to reset the autocommit as true before release to the connection pool.
conn.setAutoCommit(autoCommit); should move to the finally block
Does BoneCP (or any other pool) close connection's statements when connection is returned to pool? As I understand, it does not call actual connection's close method, so there is no automatic statement closing. So, does it close statements in any other way or do I need to close them manually?
The JDBC spec is very unclear on what should happen under normal connection close so, irrespective of the pool you use, you should always make sure to close off the statements manually. Consider what would happen to your application if you opt to switch to a different pool in the future that does not do what you expect it to do for you.
As regards BoneCP, the answer is no, it will not close off your statements for you though it can be configured to close off your connections if you forget. This is for performance reasons since some JDBC drivers will close off any still active statements internally if you close off the connection.
However, BoneCP will close off any cached statements if you have statements caching enabled.
EDIT: As of v0.8.0, support has been added to close off unclosed statements (+ print out stack trace of location where statement was opened if you want).
BoneCP (0.8.0 -RC3), there are 2 possible results,
close off with some configuration for non-cached statement only
non-close off no matter how you configure it for cached statement even you invoke the statement.close() explicitly.
There is a StatementCache class to cache the preparedStatement & callableStatement. The default is disabled. You need call BoneCPConfig.setStatementsCacheSize() with the >0 parameter to enable it. After enable the cache,
1 BoneCP.Statement.Close() will bypass the underlying statement close if it is cached.
public void close() throws SQLException {
this.connectionHandle.untrackStatement(this);
this.logicallyClosed.set(true);
if (this.logStatementsEnabled){
this.logParams.clear();
this.batchSQL = new StringBuilder();
}
if (this.cache == null || !this.inCache){ // no cache = throw it away right now
this.internalStatement.close();
}
}
2 BoneCP.Connection.close()
Will just simply clear the cache through the function "clearStatementCaches()"
The good news is MYSQL JDBC driver, Connector/J, will close all the opened statements when you close the connection through the function "closeAllOpenStatements()"
CallableStatement and PreparedStatements are precompiled. Are they done with respect to a connection? I mean, lets assume there are some 100 connection objects residing in a connection pool of an app server. There's a class that uses Callable and PreparedStatements. Lets say the method that is used for that is :
public void invokePreparedAndCallableStatements(){
//Fetches connection from pool
Connection con = getConnectionFromPool();
CallableStatement cs = con.prepareCall(.....);
cs.register...(...);
cs.execute();
...
...
PreparedStatement st = con.prepareStatement(...);
st.setXXX(..);
st.executeUpdate();
...
}
Now when the method is called for the first time, a connection is fetched from pool and the request is processed. The Callable and Prepared Statements are compiled. When the method is called another 99 times, each time a different connection is fetched from the pool, then - will the statements be complied for each connection ?
What will be the most optimal way to use statements in this context ? I can't make them (con.prepareCall() or con.prepareStatement()) static because connection isn't static.
The code is actually compiled and stored in the shared pool of the database. Any number of connections using that same code will benefit from the cache. The compiled code is kept as long as the memory limits allow.
The statements will be precompiled. Pooling will be based on your specified parameters.
Note: If you are using JDBC 3.0, you can also pool your PreparedStatements. Reference: What's new in JDBC 3.0
We have our JBoss and Oracle on separate servers. The connections seem to be dropped and is causing issues with JBoss. How can I have the JBoss reconnect to Oracle if the connection is bad while we figure out why the connections are being dropped in the first place?
Whilst you can use the old "select 1 from dual" trick, the downside with this is that it issues an extra query each and every time you borrow a connection from the pool. For high volumes, this is wasteful.
JBoss provides a special connection validator which should be used for Oracle:
<valid-connection-checker-class-name>
org.jboss.resource.adapter.jdbc.vendor.OracleValidConnectionChecker
</valid-connection-checker-class-name>
This makes use of the proprietary ping() method on the Oracle JDBC Connection class, and uses the driver's underlying networking code to determine if the connection is still alive.
However, it's still wasteful to run this each and every time a connection is borrowed, so you may want to use the facility where a background thread checks the connections in the pool, and silently discards the dead ones. This is much more efficient, but means that if the connections do go dead, any attempt to use them before the background thread runs its check will fail.
See the wiki docs for how to configure the background checking (look for background-validation-millis).
There is usually a configuration option on the pool to enable a validation query to be executed on borrow. If the validation query executes successfully, the pool will return that connection. If the query does not execute successfully, the pool will create a new connection.
The JBoss Wiki documents the various attributes of the pool.
<check-valid-connection-sql>select 1 from dual</check-valid-connection-sql>
Seems like it should do the trick.
Not enough rep for a comment, so it's in a form of an answer. The 'Select 1 from dual' and skaffman's org.jboss.resource.adapter.jdbc.vendor.OracleValidConnectionChecker method are equivalent , although the connection check does provide a level of abstraction. We had to decompile the oracle jdbc drivers for a troubleshooting exercise and Oracle's internal implementation of the ping is to perform a 'Select 'x' from dual'. Natch.
JBoss provides 2 ways to Validate connection:
- Ping based AND
- Query based
You can use as per requirement. This is scheduled by separate thread as per duration defined in datasource configuration file.
<background-validation>true</background-validation> <background-validation-minutes>1</background-validation-minutes>
Some time if you are not having right oracle driver at Jboss, you may get classcast or related error and for that connection may start dropout from connection pool. You can try creating your own ConnectionValidator class by implementing org.jboss.resource.adapter.jdbc.ValidConnectionChecker interface. This interface provides only single method 'isValidConnection()' and expecting 'NULL' in return for valid connection.
Ex:
public class OracleValidConnectionChecker implements ValidConnectionChecker, Serializable {
private Method ping;
// The timeout (apparently the timeout is ignored?)
private static Object[] params = new Object[] { new Integer(5000) };
public SQLException isValidConnection(Connection c) {
try {
Integer status = (Integer) ping.invoke(c, params);
if (status.intValue() < 0) {
return new SQLException("pingDatabase failed status=" + status);
}
}
catch (Exception e) {
log.warn("Unexpected error in pingDatabase", e);
}
// OK
return null;
}
}
A little update to #skaffman's answer. In JBoss 7 you have to use "class-name" attribute when setting valid connection checker and also package is different:
<valid-connection-checker class-name="org.jboss.jca.adapters.jdbc.extensions.oracle.OracleValidConnectionChecker" />
We've recently had some floating request handling failures caused by orphaned oracle DBMS_LOCK session locks that retained indefinitely in client-side connection pool.
So here is a solution that forces session expiry in 30 minutes but doesn't affect application's operation:
<check-valid-connection-sql>select case when 30/60/24 > sysdate-LOGON_TIME then 1 else 1/0 end
from V$SESSION where AUDSID = userenv('SESSIONID')</check-valid-connection-sql>
This may involve some slow down in process of obtaining connections from pool. Make sure to test this under load.