During architecture, design, and implementation, you must carefully consider using Java Message Service's (JMS) various transaction options. For a simple approach, you wouldn't use transactions; instead you would use acknowledgement modes, such as automatic, duplicates okay, and client. Another option uses transacted session for sending and receiving messages. Message-driven beans (MDBs) with container-managed transaction demarcation (CMTD) or with bean-managed transaction demarcation (BMTD) provides yet another alternative. This article clearly presents the available transaction options and carefully considers these options' impact on message redelivery. You'll gain a better knowledge of using JMS efficiently and effectively. (3,000 words; March 15, 2002)
Java developers must often determine whether JMS (Java Message Service) meets their requirements as a message-processing layer for their distributed systems. Issues such as application server performance, data distribution, security, and error handling must be considered. Then critical decisions about design and implementation need to be made. In this article, Thomas Laramee presents a quick overview of JMS, followed by a description of two real-world distributed systems, and analyzes potential JMS pros and cons for each system. Although the systems are architecturally similar, Thomas shows one system for which JMS is an appropriate choice and one in which it is not. He also suggests an alternative solution and offers general recommendations for when to use JMS and when to apply another strategy. (2,800 words; October 25, 2002)
Many enterprises have a federated set of homegrown and purchased applications that store their own copies of user contact information. In this age of mergers and acquisitions, a company's multiple business units or departments often carry redundant user information. Although most applications only store information about a specific application's users, the same user information is frequently stored in multiple applications. To maintain consistency and avoid customer confusion, these copies should be regularly synchronized, and they can be a challenge to maintain. In this article, we describe a simple solution to this common problem of maintaining consistent user information across various business applications. (2,500 words; March 14, 2003)
Many enterprises have a federated set of homegrown and purchased applications that store their own copies of user contact information. In this age of mergers and acquisitions, a company's multiple business units or departments often carry redundant user information. Although most applications only store information about a specific application's users, the same user information is frequently stored in multiple applications. To maintain consistency and avoid customer confusion, these copies should be regularly synchronized, and they can be a challenge to maintain. In this article, we describe a simple solution to this common problem of maintaining consistent user information across various business applications. (2,500 words; March 14, 2003)
During architecture, design, and implementation, you must carefully consider using Java Message Service's (JMS) various transaction options. For a simple approach, you wouldn't use transactions; instead you would use acknowledgement modes, such as automatic, duplicates okay, and client. Another option uses transacted session for sending and receiving messages. Message-driven beans (MDBs) with container-managed transaction demarcation (CMTD) or with bean-managed transaction demarcation (BMTD) provides yet another alternative. This article clearly presents the available transaction options and carefully considers these options' impact on message redelivery. You'll gain a better knowledge of using JMS efficiently and effectively. (3,000 words; March 15, 2002)
