From - Mon Nov 20 13:17:35 2000 X-UIDL: ad8e0cef471d0000 X-Mozilla-Status: 0011 X-Mozilla-Status2: 00000000 Received: by mailer.csit.fsu.edu (mbox gcfpc) (with Cubic Circle's cucipop (v1.31 1998/05/13) Mon Nov 20 13:03:16 2000) X-From_: fox@mailer.csit.fsu.edu Mon Nov 20 09:33:53 2000 Return-Path: Delivered-To: gcfpc@csit.fsu.edu Received: from dirac.csit.fsu.edu (dirac.csit.fsu.edu [144.174.128.44]) by mailer.csit.fsu.edu (Postfix) with ESMTP id 6E43823A07 for ; Mon, 20 Nov 2000 09:33:53 -0500 (EST) Received: from localhost by dirac.csit.fsu.edu (AIX4.2/UCB 8.7) id JAA144206; Mon, 20 Nov 2000 09:33:53 -0500 (EST) Resent-Message-Id: <200011201433.JAA144206@dirac.csit.fsu.edu> Delivered-To: fox@csit.fsu.edu Received: from burdell.cc.gatech.edu (burdell.cc.gatech.edu [130.207.3.207]) by mailer.csit.fsu.edu (Postfix) with ESMTP id 7DE5323A11 for ; Fri, 17 Nov 2000 11:02:02 -0500 (EST) Received: from tuomotu.cc.gatech.edu (tuomotu.cc.gatech.edu [130.207.5.15]) by burdell.cc.gatech.edu (8.9.1/8.9.1) with ESMTP id LAA25927 for ; Fri, 17 Nov 2000 11:02:01 -0500 (EST) Received: (from beth@localhost) by tuomotu.cc.gatech.edu (8.9.1/8.9.1) id LAA15973 for fox@csit.fsu.edu; Fri, 17 Nov 2000 11:01:57 -0500 (EST) From: beth@cc.gatech.edu (Beth Plale) Message-Id: <200011171601.LAA15973@tuomotu.cc.gatech.edu> Subject: Re: Request to Referee Paper To: fox@csit.fsu.edu Date: Fri, 17 Nov 2000 11:01:56 -0500 (EST) In-Reply-To: <39D8A82D.39FE69F6@csit.fsu.edu> from "Geoffrey Fox" at Oct 02, 2000 11:22:21 AM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Resent-To: Geoffrey Fox Resent-Date: Mon, 20 Nov 2000 09:33:53 -0500 Resent-From: Geoffrey Fox Dear Geoffrey, Here is my review of the C476 paper. Thanks for the opportunity to review; I'd be happy to do it again if you need help. best regards beth > I thought you might be able to provide me a referee report by November 17 on the paper > > C476: Effective Multicast Programming in Large Scale Distributed Systems:The DACE Approach > The PDF Text of the paper can be found at the URL > http://aspen.csit.fsu.edu/CCPEwebresource/C476eugster/C476paper.pdf > > > Thank you > Geoffrey -- Beth Plale beth@cc.gatech.edu Georgia Institute of Technology 1 404 894 1540 College of Computing Fax 1 404 894 9442 Atlanta, GA 30332-0280 U.S. --------------------------------------------------------------------------- C: Paper and Referee Metadata * Paper Number Cnnn: C476 * Date: November 17, 2000 * Paper Title: Effective Multicast Programming in Large Scale Distributed Systems:The DACE Approach * Author(s): Romain Boichat, Patrick Eugster, Rachid Guerraoui, and Joe Sventek * Referee: Beth Plale * Address: College of Computing Georgia Institute of Technology Atlanta, GA 30332-0280 Referee Recommendations. Please indicate overall recommendations here, and details in following sections. 1. publish as is 2. accepted provided changes suggested are made `accept provided ...' practical uses can be demonstrated and sufficient measurements can be made. 3. reject D: Referee Comments (For Editor Only) The paper has an interesting approach to publish/subscribe middleware. The conceptualization of communication as a hierarchy of classes of 'collections' is novel. Topic membership in the presence of failures is well thought out. I have two major reservations with the paper. First, practical experience with the system is not evident from the paper. This is to the extent that not a single application is even mentioned. Second, the measurements section is weak. The authors essentially measure only the algorithm they have developed to minimize number of messages sent in multicast settings. Though this number is convincing, I feel the strengths of the system are in its support for failures and network partitions, and in its hierarchical set of classes implementing variations of publish/subscribe. Numbers supporting the cost of these features are absent. In conclusion, I am not convinced that the system is mature enough to have been used in any practical experience setting. My recommendation is `accept provided ...' practical uses can be demonstrated and sufficient measurements can be made. E: Referee Comments (For Author and Editor) DACE is a middleware system implementing the publish/subscribe model of interaction in a distributed environment. It is topic-based, in that users subscribe to topics which presumably consist of multiple event types. Topics are organized hierarchically, in that subtopics can be derived for which a subscriber can subscribe in addition to or exclusive to the topic itself. The paper has an interesting approach to publish/subscribe middleware. The conceptualization of communication as a hierarchy of classes of 'collections' is novel. Topic membership in the presence of failures is well thought out. A strength of DACE is its tolerance to network partitions and crash failures. Topic knowledge is maintained at each site. When a network partitions, participants will renegotiate a topic member set. Crash failure is achieved by providing each participant with access to a local failure detector module which outputs hints about the closed channels with other participants. Topic member set is then renegotiated. A second strength of DACE is the conceptualization of publish/subscribe communication as a collection. For instance, the API allows 'pull' style communication by registering a callback to 'remove' an event from the collection. The notion of a collection lends itself easily to supporting collection subtypes that impose order on the events and QoS features (i.e., reliability) on the communication infrastructure. I have two major reservations with the paper. First, practical experience with the system is not evident from the paper. Second, the measurements section, primarily by the absence of results, does not convince the reviewer that the system is mature enough to have been used in any practical experience setting. Specifically, the paper contains one measurement comparing the collection subtype with the least overhead (i.e., DAStrongBag) against an unreliable multicast protocol. The measurement convincingly demonstrates the effectiveness of the first-participant algorithm developed by DACE to reduce the number of messages sent, but details were absent. I would have liked to have seen a description of the model of communication employed in the experiment, an indication of the loss rate for the unreliable multicast protocol, details as to the unreliable multicast protocol used, and a breakdown in number of events sent in the DACE case. Regarding the latter, are all events user level events or is `first-participant' traffic included? I would also have liked to have seen results supporting the two major strengths of the approach: multiple subtypes and failure recovery. What cost is associated with semantics like 'at-least-once FIFO'? What is the overhead of topic network knowledge propagation? F: Presentation Changes Discussed above in (D).