Cluster and ConnectionPooling

Question

Hi,
I need your help.
I have a WLS-Cluster with 2 instances and my webApp uses connection pooling.
By looking up for connections in the pool the application get one, but it
will never be released.
In my opinion, the webApp get the connection and looses the reference to
release it.
How do I have to configure my cluster to work well with that stuff ?
Thanks a lot,
Markus.

Answer 1

What if the code between db.connect() and db.disconnect() throws an
exception? In that case the disconnect() would never happen and you
have basically lost a connection from your pool. Could it be that in the
clustered version the code between those calls is throwing an exception
that the non-clustered version is not being thrown??
Regards
alex
"M. Hammer" wrote:
Here it is...
But I don't think that you'll see anything in the source.
My WebApp works well standalone, but not in a CLUSTER.
By the way I put an new WLS Instance to a standalone running one,
connections will be used and never be released.
Exactly the connection will be open up in the partner-instance of the
requested one.
So, I think that the cluster make the connection-pool-request to both,
Instance 1 and 2, then on one connection, the reference is lost.
Hope you know what I mean. ;o)
public class MyServlet extends HttpServlet
public void doGet (HttpServletRequest request, HttpServletResponse
response)
throws IOException, ServletException
Db db = mySession.getDb();
if( mySession.useConnectionPooling && !mySession.useStaticDb ) {
db.connect();
if( mySession.useConnectionPooling && !mySession.useStaticDb ) {
db.disconnect();
"Michael Reiche" <[email protected]> schrieb im Newsbeitrag
news:[email protected]...
Ok, that was the code that connects. Now the code that uses it and
disconnects.
Mike
"M. Hammer" <[email protected]> wrote in message
news:[email protected]...
Ok, here it is...
public boolean connect()
Hashtable jndi_env = new Hashtable();
Context context = null;
DataSource ds = null;
if ( driverRegistered )
if( isConnected() ) {
disconnect();
if( useConnectionPooling ) {
try {
if( activateSettings ) {
//--- JNDI-Settings aus App-Property-Datei
setzen
jndi_env.put( Context.INITIAL_CONTEXT_FACTORY,
JNDI_factory_initial );
jndi_env.put( Context.PROVIDER_URL,
JNDI_provider_url );
jndi_env.put( Context.SECURITY_PRINCIPAL,
JNDI_security_principal );
jndi_env.put( Context.SECURITY_CREDENTIALS,
JNDI_security_credentials );
if( activateSettings ) {
context = new InitialContext( jndi_env );
} else {
context = new InitialContext();
ds = (DataSource)context.lookup(
myDataSource );
connection = ds.getConnection();
} catch( SQLException ex_sql ) {
return false;
} catch( NamingException ex_nam ) {
return false;
} else {
try {
connection = DriverManager.getConnection(url, login,
password);
} catch(SQLException ex) {
return false;
else
return false;
return true;
"Michael Reiche" <[email protected]> schrieb im Newsbeitrag
news:[email protected]...
Can you post the code that gets the connection, uses it, and returns
it
to
the pool?
Mike
"M. Hammer" <[email protected]> wrote in message
news:[email protected]...
Hi,
I need your help.
I have a WLS-Cluster with 2 instances and my webApp uses connectionpooling.
By looking up for connections in the pool the application get one,
but
it
will never be released.
In my opinion, the webApp get the connection and looses the
reference
to
release it.
How do I have to configure my cluster to work well with that stuff ?
Thanks a lot,
Markus.