Weblogic datasource issue

Question

I have a datasource targeted to a cluster with two weblogic managed servers i.e 239.112.112.47:8001 and 239.112.112.48:9001

The datasource has deafult max capacity for the connection pool set to 15.

Thus, I am expection a total of 15*2=30 connections.

However, my client code gets only 15 connections that too on a single server(239.112.112.47:8001) and not on both servers of my cluster.

Below is my code:

public static void main(String ar[])throws Exception
{
InitialContext ic=null;
try{
Hashtable env = new Hashtable();
env.put(Context.INITIAL_CONTEXT_FACTORY, JNDI_FACTORY);
env.put(Context.PROVIDER_URL, 239.112.112.47:8001);
ic = new InitialContext(env);
}
catch(Exception e){
System.out.println("\n\n\t Unable To Get The InitialContext => "+e);
}
Connection con[]=new Connection[100];
try{

for(int i=0;i<100;i++)
{
   DataSource ds=(DataSource)ic.lookup("MyDS");   //Enter your DataSource or MultidataSource JNDI name Here
   con[i]=ds.getConnection();
   System.out.println("\n\n\t GOT CONNECTION con["+i+"]: "+con[i]);
   Thread.sleep(1000);
}

Do you mean that all of your client requests go through the same managed server? In this case, the maximum number of 15 connections seems OK to me and your problem is not with the datasource, but with the cluster setup. And yes, your assumption is correct, the default max capacity is a setting that is applied to each server within the cluster. — Viccari
Yes, all the client requests are going to a single server i.e 239.112.112.47:8001. What do i do to get 30 JDBC connections ? I have setup the cluster with all default setting i.e. I have not made any explicit changes. — shifahim
You need to figure out why your requests are being redirected to a single server. Once you are able to balance them to both managed servers, your connection count will probably reach 15 per server. Your data source setup seems OK. It seems to me that your issue is more related to the balancing of the handling of your client requests (on the cluster) than on the data source itself. — Viccari
Are you sure that in the client code even if we specify one server i.e. env.put(Context.PROVIDER_URL, 239.112.112.47:8001); the client request should ideally be going to the cluster to both the managed servers ? If the client code is incorrect then please help identify the correct code that takes the request to both the servers in the cluster. — shifahim
Unfortunately, I am not experienced with WebLogic clustering. I suggest you to make another question with this subject. — Viccari

Musannif Zahir Musannif Zahir · Accepted Answer · 2012-12-20T12:10:07

It sounds like you have one of these issues:

Your datasource is targeting a single managed server instead of the cluster
Your cluster has a single member instead of both of the managed servers
Your load balancer is directing traffic to a single node within the cluster

Weblogic datasource issue

1 Answers