Planning installation topologies

Servers can be configured in different topologies, and the installation procedure you use varies according to the type of topology you select. To help you decide which installation method is best for you, four common topologies are described, along with diagrams and detailed steps to guide you through the installation procedure.

Before you begin

WebSphere Process Server includes server software and client software. You must install the server software, but the client software is optional. Typically, you install the Message Service Client software on systems where you have C, C++, or .Net messaging applications or C++ Web services running.

Establishing the server environment for WebSphere Process Server involves:
  1. Installing a set of core system files called system files. These system files include command files and other product binary files. System files can be updated by installing refresh packs or fix packs.
  2. Configuring either a stand-alone profile or a number of managed profiles (in a network deployment environment). Each profile provides a separate runtime environment for the server processes that are defined to it. During management and running of these servers, the configuration files, data files, and log files can be created, read, updated, or deleted, but there is only read access to the system files. When you use a refresh pack to update a set of system files, then all of the profiles associated with the installation will begin using the updated files as they are started.
Figure 1. Separation of system files and profiles
Separation of system files and profiles

You can install the WebSphere Process Server system files once and then create multiple profiles on the same system. Or, you can install multiple separate versions of the product, each with its own profiles, on one or more systems. The product files are installed in the installation root (WebSphere Application Server_HOME). In some environments, the distinction is made between the directory for the core product files (WebSphere Application Server_INSTALL_ROOT) and the directory for the profile files (USER_INSTALL_ROOT).

WebSphere Process Server is built on WebSphere Application Server Network Deployment, version 6. You can augment WebSphere Application Server profiles to become WebSphere Process Server profiles, converting from a WebSphere Application Server scenario to WebSphere Process Server, for deployment manager and stand-alone profiles; custom profiles can be converted only if they have not been federated.

You begin the installation for each topology by performing either a Complete or a Custom installation (by selecting Complete or Custom on the Installation type panel in the installation wizard). A Complete installation installs the system files and automatically creates a default stand-alone server profile with a samples gallery. A Custom installation installs the core product files but does not create a stand-alone server profile.
Important: This simple installation, described in "Topology 1" below, is adequate for evaluation or proof-of-concept purposes related to the programming model. "Topology 4," described later in this topic, is adequate for evaluation or proof-of-concept purposes related to non-functional requirements such as availability or scalability.

You can then customize your system in different ways as described in the following topologies.

Types of installations

You can install the server on server hardware as individual stand-alone servers or as a managed group of servers.
  • A stand-alone server profile has its own administrative console and all sample applications (if you performed a Complete installation or selected to install the sample applications gallery feature during a Custom installation). Each stand-alone server is fully operational and is managed independently from all other servers.

    There are two main types of stand-alone server installations:

    • One stand-alone server on one system (described in "Topology 1")
    • Multiple stand-alone servers on a single installation on one system (described in "Topology 2")
  • A managed group of servers, also known as a cell, uses a deployment manager for centralized administration tasks, such as managing the configuration for all of the managed nodes in its cell and deploying applications to selected managed nodes in the cell. All profiles in the cell share command files and other product binary files. The main reason to use managed nodes in a cell rather than using the same number of stand-alone servers is the centralized administration that the deployment manager provides for the cell.

    If you want to balance workload, such as service requests, over a set of servers, you can create a server cluster, then add servers as members of that cluster. You can also create a backup cluster, to provide failover support for the server cluster to which it is assigned.

    There are two main types of server cell installations:

    • Deployment manager and managed nodes (cell) on a single installation on one system (described in "Topology 3")
    • Deployment manager and managed nodes on a single installation on multiple systems, with a server cluster (described in "Topology 4")

Steps for this task

Result

After completing the installation, you will have the appropriate number of profiles and servers for your topology.

What to do next

You can run the samples to explore their functions (if you installed the Sample applications gallery feature). You can also deploy your own mediation modules, or adapt the server and bus environment to your needs.


Last updated: Wed 01 Nov 2006 07:47:12

(c) Copyright IBM Corporation 2005, 2006.
This information center is powered by Eclipse technology (http://www.eclipse.org)