JMS Compliance in the Narada Event Brokering System.

作者: Shrideep Pallickara , Geoffrey C. Fox

DOI:

关键词: Asynchronous communicationRemote procedure callComputer networkComputer scienceJava Message ServiceDistributed Component Object ModelJavaPeer-to-peerCommon Object Request Broker ArchitectureWorld Wide WebSun Microsystems

摘要: Narada is a distributed event brokering system based on the publish/subscribe paradigm and designed to run very large network of broker nodes. This paper describes process achieving Java Message Service (JMS) compliance in system. The also our strategy for bringing Narada’s benefits – scaling, availability, routing efficiencies failure resiliency existing JMS compliant systems. Finally, we include performance data from Messaging systems queuing products such as IBM's MQSeries [1] Microsoft's MSMQ [2]. model with their store-and-forward mechanisms come into play when sender message expects someone handle while imposing asynchronous communication guaranteed delivery constraints. A widely used standard messaging Passing Interface Standard (MPI) [3]. MPI high both massively parallel machines workstation clusters. frameworks classical remote procedure calls CORBA [4] OMG, DCOM [5] Microsoft RMI [6] Sun Microsystems. In messages publisher subscriber within purview oriented middleware (MOM), which responsible right content producer consumers. Industrial strength solutions domain like TIB/Rendezvous [20] TIBCO SmartSockets [19] Talarian. Other related efforts research community Gryphon [21], Elvin [22] Sienna [23]. push by publish subscribe features its JINI [13] [14]. One goals offer unified API across implementations. Various implementations SonicMQ [15] Progress, JMQ [16] iPlanet FioranoMQ [17] Fiorano. [8,9,10,11,12] cluster architecture, results creation “small world networks” [7], allows support heterogeneous client configurations that scale arbitrary size. Protocols controlling addition nodes ensure brokers being added do not result bandwidth degradation because interconnectivity. Changes addition/removal channels between brokers. incorporates scheme controlled propagation information pertaining changes relevant sections network. tandem organization makes efficient maps (BNM) an one. paths computed, reach any given destination, these BNMs tend be since protocols associated connections have ensured degradations. Clients interested using can attach themselves one specify interest type events they are service routes events, satisfy constraints specified clients. prolonged disconnects roam (in response suspicions or better times) other node Events published during intervening period, roams, must still delivered clients originally had events. satisfied presence connection failures. provides hierarchical dissemination calculation disseminations, BNMs, ensuring computed comprise only those failed been suspected. Community Grids [27] needs encompass peer (P2P) [24] traditional centralized middle tier style interactions. We base P2P interactions through JXTA [19], set open, generalized communications. Details integration found [26]. strategies bring functionality organized follows. section 1 describe what implies. Section 2 rationale Narada, 3 so. 4 applications were test compliance. 5 presents help leverage settings guarantees accorded conformance specification. 6

参考文章(8)
Jim Waldo, Ken Arnold, The Jini Specifications Addison-Wesley Longman Publishing Co., Inc.. ,(2000)
Shrideep Bhaskaran Pallickara, Geoffrey C. Fox, A grid event service Syracuse University. ,(2001)
Shrideep Pallickara, Geoffrey Fox, The Narada Event Brokering System: Overview and Extensions parallel and distributed processing techniques and applications. pp. 353- 359 ,(2002)
Geoffrey C. Fox, Shrideep Pallickara, An event service to support Grid computational environments Concurrency and Computation: Practice and Experience. ,vol. 14, pp. 1097- 1127 ,(2002) , 10.1002/CPE.689
Marcos K. Aguilera, Robert E. Strom, Daniel C. Sturman, Mark Astley, Tushar D. Chandra, Matching events in a content-based subscription system principles of distributed computing. pp. 53- 61 ,(1999) , 10.1145/301308.301326
Duncan J. Watts, Steven H. Strogatz, Collective dynamics of small-world networks Nature. ,vol. 393, pp. 440- 442 ,(1998) , 10.1038/30918
Antonio Carzaniga, David S. Rosenblum, Alexander L. Wolf, Achieving scalability and expressiveness in an Internet-scale event notification service principles of distributed computing. pp. 219- 227 ,(2000) , 10.1145/343477.343622
Ozgur Balsoy, Ahmet Uyar, Shrideep Pallickara, Aleksander Slominski, Geoffrey Fox, Dennis Gannon, Community Grids international conference on computational science. pp. 22- 38 ,(2002)