Webogic cluster query

Question

Hi,
We are implementing our application on weblogic cluster. Earlier it was on single instance. Had some question's regarding WLcluster working.
1> Do we need separate path for application files or can it be at shared location used different managed servers.
2> If there are two app server say server1 (ip 10.10.10.10 port 7002) and server2 (ip 10.10.10.11 port 7002) in single cluster, where should we hit the request and how the load is equally balanced among both server.
3> It is better to create different resources (jdbc, jms) for each server in a cluster or should we create single and assign it to whole cluster (what is best practice).
Thanks in advance..

Answer 1

1> Do we need separate path for application files or can it be at shared location used different managed servers.
NO. Its Not Required. Basically, you login into weblogic admin console, and from Deployments section you will Deploy your App to the Cluster. So at that time, when you browse and choose your app ear file, weblogic will store it internally in one of the folder structures under your domain root folder. Usually, it goes to like domain/servers/adminserver/upload folder structure. Or it can be anything. But the point is, you do not need to specify any particular location. Some people what they do is, upload the ear file using like ftp client to a folder on the machine having admin server. And at time of deployment, they choose this location. You do not need to worry about how this gets shared across another managed server on a different host. Weblogic will take care of this.
2> If there are two app server say server1 (ip 10.10.10.10 port 7002) and server2 (ip 10.10.10.11 port 7002) in single cluster, where should we hit the request and how the load is equally balanced among both server.
NOW, this is where you need to configure something called Web Server or Load Balancer in front of the cluster. Many customers use Apache HTTP Server as Web Server and configure one file like httpd.conf file to give above details. Now end users will access this Apache host and port like apachehost and port 80. Apache internally will do Load Balancing, fail over etc. So you need to configure a load balancer in front. Generally, we do not give individual host, port to access the application on cluster. Refer online docs for more details.
3> It is better to create different resources (jdbc, jms) for each server in a cluster or should we create single and assign it to whole cluster (what is best practice).
Assign Single and Target to the Entire Cluster. Say create a DataSource, give all db connection pools, and Target this to Entire Cluster (internally, it will replicate and accessible to both managed servers). Now for JMS Queues, I guess you do specific to each managed server with some prefixes like _1, _2 etc. I do not have all details. But general practice is to target to cluster.
Thanks
Ravi Jegga