Platform Framework
R24 AMR | Min(s) read

Configuring WebLogic Script (cluster) for Temenos Transact Java Deployment

Similar to single server mode, you can use scripts to run a multi-environment. To use the scripts you need to setup a cluster called TAFJCluster.

To launch a COB, one of the following types of URL is used. If in a cluster mode, then you need to launch the COB through multiple URLs, as shown below.

http://10.44.245.223:9548/TAFJEE/Execute?command=START.TSM

http://10.44.245.224:9548/TAFJEE/Execute?command=START.TSM

This requires a special setup, which is explained below.

Setting up Distributed Queues in Script Mode

To setup distributed queues in script mode,

  1. Run the script below.

    wlst.sh -loadProperties %TAFJ_HOME%\appserver\weblogic\tafj.properties %TAFJ_HOME%\appserver\weblogic\TAFJJMS-Distributed.py

  2. Run the script below.

    wlst.sh -loadProperties %TAFJ_HOME%\appserver\weblogic\tafj.properties %TAFJ_HOME%\appserver\weblogic\TAFJJMSQUEUE-Distributed.py

    This creates distributed Round-Robin queues in the cluster for each server but not in the admin server.

Setting up Distributed Queues in Console Mode

To setup distributed queues in console mode,

  1. Create a new Queue and choose Distributed Queue on the following page.
  2. Specify a Name and JNDI Name. Click Next.
  3. Click Advanced Targeting and choose the Subdeployment setup. Click Finish.

Setting up a Grind Link Datasource

The procedure for setting up a Grind Link Datasource (a datasource that targets an Oracle RAC Cluster) is similar to that of running through a single datasource wizard. In this case, you need to select the Fan Enabled check box and add the host or ports of each node in the RAC.

Below is an example of the URL that you need to specify where TAFJWL is the ORACLE_SID and the hosts or ports are highlighted.

jdbc:oracle:thin:@(DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCP)(HOST=10.44.245.186)(PORT=1521))(ADDRESS=(PROTOCOL=TCP)(HOST=10.44.245.187 )(PORT=1521)))(CONNECT_DATA=(SERVICE_NAME=TAFJWL)))

Setting up each Node so that TAFJEE.ear can be Deployed to the Cluster

When you are using multi-node mode, each node must be aware of its classpath. If TAFJ is shared over a network share, then the classpath can be shared using the same directory structure. Otherwise, if TAFJ is installed separately on each node, the directory structures might not match. 

There are different ways to achieve this. The most common are presented below. The recommended one is to follow same mechanism than the one applied for a single server by using script configuration. For more information, see the reference URLs shown in the sections below.

Load Balancing

Switch off Server Affinity to load balance across distributed queues for transactional queues such as OFSQueue and OFSReplyQueue. Do not perform this activity for t24EXECQueue. For more information, see Setting up t24ExecQueue for multi-node.

Server Affinity Enabled - Specifies whether a server instance that is load balancing consumers or producers across multiple members destinations of a distributed destination, will first attempt to load balance across any other physical destinations that are also running on the same server instance.

MBean Attribute (Does not apply to application modules): LoadBalancingParamsBean.ServerAffinityEnabled

Setting up t24ExecQueue for Multi-node

This section creates a new connection factory for t24ExecQueue called t24ConnectionFactory.

Take the defaults and run through the wizard without going to Advanced Targeting (use the subdeployment).

Once the connection factory is created, uncheck the Load Balancing Enabled check box as each message will be targeted to each node.

Copyright © 2020- Temenos Headquarters SA

Published on :
Monday, May 27, 2024 5:10:43 PM IST