Build a Wadi cluster environment for WebSphere application Server Community Edition V2.1

Since the WebSphere application Server Community Edition (is CE) version 2.1, was CE has added support for the WADI cluster beyond the Tomcat native cluster. Compared to the Tomcat native cluster, the WADI cluster also provides session replication between cluster members to avoid single point failures and disaster recovery, while leveraging load balancing to increase application availability. Additionally, in a was CE cluster environment with farming deployed, all nodes within the cluster respond to these requests synchronously if you are deploying or lifecycle management on a node. Farming this feature makes application management (such as deployment, startup, shutdown, and uninstall) more logical AND transparent: once deployed, applications can be deployed to each node in the was CE server cluster; Again, just once, You can manage the lifecycle of the application on all nodes in the cluster. This article introduces the was CE WADI cluster and the farming plug-in from both the implementation mechanism and the sample configuration

Introduction

WebSphere Application Server Community Edition (hereafter referred to as CE) is a fully compliant Java Platform, Enterprise Edition 5 (Java EE 5) specification, certified application Server. That is, being CE contains all the components that support Java EE 5 implementations: Web containers, EJB containers, messaging services, command-line management, and so on, to develop and run Java EE applications. Although the Java EE 5 specification does not provide specification requirements for cluster implementations, in a real-world application environment, clustering is widely used as a common solution to provide extensibility and ensure high availability of applications. To implement clustering, it is common practice to ensure high reliability of applications by extending multiple server nodes so that session and node information is replicated within the cluster. But as the nodes in the cluster increase, how to ensure the efficiency of the cluster and application availability? Especially in heterogeneous application server environment, is there any more ideal solution? In addition, in order to achieve the unity of each node in the cluster, each node needs to configure the same application, what is a better way to do the bulk deployment, so as to improve the efficiency of cluster management? This paper, by introducing the integrated WADI module and farming module in the was CE server, answers the questions mentioned above, and demonstrates how to use the WADI and farming modules to build the CE cluster in the was CE server environment at the end of the paper. The content mentioned in this paper is a summary of the author's study, which is only for readers ' reference, to help readers understand the concepts and familiar with the deployment process, the view is not representative of IBM.

The realization mechanism of WADI cluster

WADI's full name is the Web application distribution infrastructure, the Web application distributed architecture, which, as we can see, is WADI this project was designed to address the Web-tier application state in a clustered environment. Acting As the project matures, coupled with WADI's own unique features and solutions, it has evolved into a more generic framework for distributed State and service management. Because WADI is a stand-alone open source project that does not depend on any existing application container, the role and significance of WADI is more pronounced in a heterogeneous environment.

Was CE from version V2.1 began, will WADI integrated into their own server distribution package, so that the users of was CE in addition to the original Tomcat cluster, can also consider the choice of WADI cluster scheme.

Introduction and Characteristics of WADI

WADI's design goals are briefly summarized as follows:

Provides pluggable, stack-style page scheduling (paging) and persistence (persistence) mechanisms for distributed environments;

Provides a cluster base with strong scalability, high availability, self segmentation (self-partitioning) and self-healing (self-healing);

A transparent migration (migration) that provides status to a request (state to invocation) or a request to a status (invocation to state);

Provides a pluggable state replication mechanism.