From nsg7@cornell.edu Tue Apr 18 10:09:39 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-1.9 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from authusersmtp.mail.cornell.edu (granite1.mail.cornell.edu [128.253.83.141]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3IE9d224521 for ; Tue, 18 Apr 2006 10:09:39 -0400 (EDT) Received: from [127.0.0.1] (cpe-24-59-77-191.twcny.res.rr.com [24.59.77.191]) (authenticated bits=0) by authusersmtp.mail.cornell.edu (8.13.1/8.12.10) with ESMTP id k3IE9boZ020133 for ; Tue, 18 Apr 2006 10:09:38 -0400 (EDT) Message-ID: <4444F35F.8020102@cornell.edu> Date: Tue, 18 Apr 2006 10:10:39 -0400 From: Nick Gerner User-Agent: Mozilla Thunderbird 1.0.6 (Windows/20050716) X-Accept-Language: en-us, en MIME-Version: 1.0 To: egs+summary@cs.cornell.edu Subject: PAPER 23 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit In "The Physiology of the Grid..." Foster, Kesselman, Nick, Tuecke present the Open Grid Services Architecture (OGSA) which seeks to provide a standard architecture for integrating resources of "virtual organizations". OGSA is meant to describe the "physiology" of the grid, defining functionality, while Globus (an open-source grid computing toolkit) is meant to describe the "anatomy" of the grid, defining protocols. These resources could be computational, network, storage, etc. The authors begin by presenting the evolution of grid services along two dimensions. The first is distribution. Today services are highly distributed across the country or the world and highly hetrogenous in scale and platform. The second is application. Grid computing was originally targeted at scientific collaboration, but today integrates between businesses and customers and between businesses. The rest of the paper presents OGSA. OGSA defines a large-scale architecture providing many services from discovery to remote invokation to authentication, authorization and auditing. OGSA does not define implementations, but instead identifies interfaces, composition of services and requires uniform semantics of similar services. OGSA defines these semantics through a set of interfaces and conventions (such as lifetime management and service handle definitions) much as the Java specification and API does. OGSA doesn't define a system or even a reference implementation (although Globus does). Instead OGSA provides an open standard for businesses and other organizations to conform to in order to interoperate with one-another. Because the scope of OGSA is so large, providing a distributed interaction protocol for very large-scale system integration, it's hard to tell it's strengths and weaknesses. Obviously there are issues that will arise, one of which is adoption. Time will tell if OGSA provides the needed standards. Ultimately the contribution of OGSA is almost a political one, rather than technical, instead relying on existing technologies (web services, HTTP, etc.). In "On Death, Taxes, and the Convergence..." Foster and Iamnitchi compare the work of the grid computing field and peer-to-peer systems. Specifically, grid is defined as being structured and reliable. Systems are assumed to be well-known (at least relatively to peer-to-peer). These systems perform a wide variety of diverse services to clients and are long-lived. Clients may aggregate arbitrary state within the grid over a long period of time. And this state may be shared with other clients who might use it to aggregate different state elsewhere within the grid. Peer-to-peer systems on the other hand are said to be application specific, providing specific short-lived services. A single peer-to-peer system might be large, but each node provides a single service, and clients are said not to aggregete intermediate state in the network for use later or by other clients. Instead peer-to-peer systems are resilient to failures, do not depend on administrative organization and nodes may be unknown to the vast majority of other nodes. Foster and Iamnitchi hope that these fields will (and argue that they must) converge on systems and techniques that combine the benefits of both fields. That is, systems which provide the rich and diverse services, aggregation of arbitrary intermediate state, and collaboration between clients of grids with the robust, resilient and self-organizing properties of peer-to-peer systems. Just as OGSA makes a meta-contribution to the field, this paper too argues for adoption and identifies opportunites for integration of work already existing. From tc99@cornell.edu Thu Apr 20 01:53:59 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-1.9 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from iago.cs.cornell.edu (iago.cs.cornell.edu [128.84.96.10]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3K5rw204905 for ; Thu, 20 Apr 2006 01:53:58 -0400 (EDT) Received: from postoffice10.mail.cornell.edu ([132.236.56.14]) by iago.cs.cornell.edu with Microsoft SMTPSVC(6.0.3790.1830); Thu, 20 Apr 2006 01:53:01 -0400 Received: from webmail.cornell.edu (hermes21.mail.cornell.edu [132.236.56.20]) by postoffice10.mail.cornell.edu (8.12.10/8.12.6) with ESMTP id k3K5r0kY015581 for ; Thu, 20 Apr 2006 01:53:00 -0400 (EDT) Received: from 24.59.114.243 by webmail.cornell.edu with HTTP; Thu, 20 Apr 2006 01:53:01 -0400 (EDT) Message-ID: <3769.24.59.114.243.1145512381.squirrel@webmail.cornell.edu> Date: Thu, 20 Apr 2006 01:53:01 -0400 (EDT) Subject: paper 23 From: "Theodore Ming Shiuan Chao" To: egs@cs.cornell.edu User-Agent: SquirrelMail/1.4.5 MIME-Version: 1.0 Content-Type: text/plain;charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Priority: 3 (Normal) Importance: Normal X-OriginalArrivalTime: 20 Apr 2006 05:53:01.0224 (UTC) FILETIME=[AEEE4A80:01C6643E] Grid is a different response to issues similar to some that P2P applications have been developed to deal with: sharing and allocating resources over widely distributed networks. The main point of the OGSA paper is an outline of the distributed services architecture. Hosts implement service factories that create transient service instances upon request. To tolerate failures that could potentially cause orphaning of service requests, OGSA uses a soft-state approach where every instance has a pre-negotiated lifetime, and extending the lifetime requires explicit requests from clients. To advertise and discover services, OGSA requires a centralized registry service. Grid and P2P both seek to optimize the same general idea: the distribution of resources. The only difference is the perceived usage of the two systems. While Grid tends to be used in more scientific settings where such things as incentives are not a concern and the number of distinct independent entities in the network is small, P2P is often associated with file-sharing services where leechers who do not share their bandwidth are common and the network has completely different characteristics. However, that is just a difference in assumptions. One could easily take, for instance, the file storage distribution system of Samsara and apply it over a network that Grid is running on. Or one could apply a resource exchange application such as SHARP. From pjk25@cornell.edu Thu Apr 20 02:46:38 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-2.1 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from authusersmtp.mail.cornell.edu (granite1.mail.cornell.edu [128.253.83.141]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3K6kb216543 for ; Thu, 20 Apr 2006 02:46:37 -0400 (EDT) Received: from [10.0.1.201] (cpe-69-207-37-155.twcny.res.rr.com [69.207.37.155]) (authenticated bits=0) by authusersmtp.mail.cornell.edu (8.13.1/8.12.10) with ESMTP id k3K6kV8U005593 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Thu, 20 Apr 2006 02:46:36 -0400 (EDT) Mime-Version: 1.0 (Apple Message framework v749.3) Content-Transfer-Encoding: 7bit Message-Id: <4F91BE3C-583A-41C9-8F76-CBEDB264BF95@cornell.edu> Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed To: egs+summary@cs.cornell.edu From: Philip Kuryloski Subject: PAPER 23 Date: Thu, 20 Apr 2006 02:46:44 -0400 X-Mailer: Apple Mail (2.749.3) OGSA: The authors wish to provide an Open Grid Services Architecture which will allow the successful commercial application of grid computing. This system will provide QoS guarantees, robustness, and flexible resource allocation, as well as integration with existing legacy systems. Aspects of the system are based on the existing Globus Grid Computing Toolkit. Other web based standards such as SOAP, WSDL, and others also used. The main functional unit generated by OGSA is a virtual organization (VO). The proposed grid service interfaces are GridService, Notification-Source, Notification-Sink, Registry, Factory, and HandleMap, and are used to compose a set of services for a VO. The following higher-level services achievable by the systm include distributed data management services, workflow services, auditing services, instrumentation and monitoring services, problem determination services for distributed computing, and security protocol mapping services. Essentially the system organizes and arbitrates the combined resources of a large number of networked computers. Every resources is represented as a service, and users generate virtual organizations which make use of collections of these services. OGSA proposes to use a number of web standards as well as the Globus toolkit to implement the system. This document is a recommendation for the final system, so there is no evaluation of an actual OGSA system. Thus, it is difficult to comment on it's merit. CONVERGENCE: The authors primary claim is that P2P and Grid computing systems address the same greater goal, yet each thus far has focused primarily on two different aspects of such a system: P2P addresses failure but not infrastructure, and Grid computing infrastructure but not failure. As an ideal system would address both issues, the authors see an eventual convergence of these two technologies. There are however, several significant differences between current P2P and Grid systems. Grid systems generally are composed of resources for which there is some level of trust, and the services offered can be specialized. P2P systems cannot assume trusted resources, and provide more general services. Grid resources are also generally more powerful and well connected, as where P2P resources are varied and may be poorly connected. These characteristics also lend to more hierarchical organization in Grid systems, and larger file transfers between nodes. Grid systems are also generally composed of much smaller numbers of nodes. The authors argue that despite this disparities, both types of systems would benefit from the advances made by the other. Grid systems will grow to encompass larger numbers of nodes, and P2P networks will benefit from more structure, for example. However, in my own opinion there seem to be some non-convergent aspects of the two systems, or rather two disparate assumptions or design philosophies associated with each system: fairness is a requirement and must be enforced in P2P, while Grids serve the same greater purpose and so fairness between nodes is not required. Based on this difference, it is not clear to me that complete convergence between the two types of systems will be achieved or will be optimal. From sh366@cornell.edu Thu Apr 20 02:52:09 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-2.4 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from penguin.cs.cornell.edu (penguin.cs.cornell.edu [128.84.96.11]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3K6q8217802 for ; Thu, 20 Apr 2006 02:52:08 -0400 (EDT) Received: from postoffice10.mail.cornell.edu ([132.236.56.14]) by penguin.cs.cornell.edu with Microsoft SMTPSVC(6.0.3790.1830); Thu, 20 Apr 2006 02:51:15 -0400 Received: from orpheus3.dataserver.cornell.edu (orpheus3.dataserver.cornell.edu [128.253.161.167]) by postoffice10.mail.cornell.edu (8.12.10/8.12.6) with ESMTP id k3K6pF0R028743 for ; Thu, 20 Apr 2006 02:51:15 -0400 (EDT) Message-ID: <985722237.1145515873353.JavaMail.webber@orpheus3.dataserver.cornell.edu> Date: Thu, 20 Apr 2006 02:51:13 -0400 (EDT) From: Huang Shiang-Jia To: egs+summary@cs.cornell.edu Subject: PAPER 23 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: uPortal WEB email client 3.0 X-OriginalArrivalTime: 20 Apr 2006 06:51:15.0427 (UTC) FILETIME=[D1A35B30:01C66446] * The Open Grid Services Architecture (OGSA) describes an architecture for a service-oriented grid computing environment. OGSA, based on several other web service technologies, aims to be largely agnostic in relation to the transport-level handling of data. * Basically, OGSA is a distributed interaction and computing architecture based around services, assuring interoperability on heterogeneous systems so that different types of resources can communicate and share information. * OGSA extends the earlier article "The Anatomy of Grid" that defines Grid technologies as supporting the resource sharing in distributed virtual organizations (VOs) in three respects. (i) It focuses on the nature of services that respond to protocol messages. Here Grid is regarded as an extensible set of Grid services, services that conform to a set of conventions and support lifetime management, discovery of characteristics and so on. (ii) It shows how Grid technologies can be aligned with web services technologies. The Web Service Description Language (WSDL) is used here to achieve self-describing services and interoperable protocols. (iii) It discusses the commercial applications rather than scientific applications of this architecture. * It is worth noting that all components in the Grid environment are virtualized. Virtualization of this model enables mapping common service semantic behavior seamlessly onto native platform facilities. * This paper compares and contrasts peer-to-peer and grid computing. It arrives at four conclusions. (i) Peer-to-peer and grid computing both are concerned with the pooling and coordinated use of distributed resources within virtual communities. (ii) Peer-to-peer and grid computing both are constructed as overlay structures that operate independently of institutional relationships. (iii) Peer-to-peer computing addresses failure but not yet the infrastructure where as grid computing addresses infrastructure but not yet the failure. (iv) Because of their commonalities and complementary strengths and weaknesses mentioned above, they are likely to grow close over time. From asg46@cornell.edu Thu Apr 20 09:09:10 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-2.3 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from iago.cs.cornell.edu (iago.cs.cornell.edu [128.84.96.10]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3KD9A215871 for ; Thu, 20 Apr 2006 09:09:10 -0400 (EDT) Received: from postoffice10.mail.cornell.edu ([132.236.56.14]) by iago.cs.cornell.edu with Microsoft SMTPSVC(6.0.3790.1830); Thu, 20 Apr 2006 09:08:44 -0400 Received: from webmail.cornell.edu (hermes21.mail.cornell.edu [132.236.56.20]) by postoffice10.mail.cornell.edu (8.12.10/8.12.6) with ESMTP id k3KD8hi4003552 for ; Thu, 20 Apr 2006 09:08:43 -0400 (EDT) Received: from 128.84.98.251 by webmail.cornell.edu with HTTP; Thu, 20 Apr 2006 09:08:43 -0400 (EDT) Message-ID: <4769.128.84.98.251.1145538523.squirrel@webmail.cornell.edu> Date: Thu, 20 Apr 2006 09:08:43 -0400 (EDT) Subject: paper 23 From: "Abhishek Santosh Gupta" To: egs+summary@cs.cornell.edu User-Agent: SquirrelMail/1.4.5 MIME-Version: 1.0 Content-Type: text/plain;charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Priority: 3 (Normal) Importance: Normal X-OriginalArrivalTime: 20 Apr 2006 13:08:45.0070 (UTC) FILETIME=[8DE09AE0:01C6647B] GRID Physiology of the Grid the authors define an Open Grid Services Architecture (OGSA) that supports the creation, management and invocation of stateful, transient services as named entries with dynamic lifetimes. Merits of this service-oriented model are as follows: All components of the environment are virtualized. Virtualization brings to the table a large number of advantages such as composition of services regardless of implementation, management of resources across heterogeneous platforms with local or remote transparency and the ability to map common service semantic behavior seamlessly onto native platform facilities. It evolves the Globus toolkit using a number of components such as "gatekeeper", Meta Directory Service, and Grid Security Infrastructure. the underlying Web Services framework supports in dynamic discovery of services and nescessiates mechanisms for registering and discovering interface definitions and endpoint implementation descriptions. different Virtual Organization (VO) structures: Simple Hosting Environment : consists of a single administrative domain and supporting native facilities for service management Virtual Hosting Environment : VOs dealing with multiple heterogeneous environments require another higher layer to delegate requests to the lower level Collective Operations : Virtual Hosting Environments which provides VO participants with more sophisticated or collective end-to-end services. each Grid Service Request results in creation of a grid service instance with some initial lifetime. the user application is required to generate "keepalive" requests to the Grid Service that it has created. the lifetime of a request can be extended by a client though granting such a request is not mandatory. the network protocol must provide reliable transport, authentication and delegation, ubiquity and the GSR format into consideration. Higher level services must include distributed data management services, workflow services, auditing services, instrumentation and monitoring services and security mapping services. P2P and Grid Computing both the systems deal with resource sharing in virtual communities. Comparisons: Grids generally deal with more sophisticated/complex services and application while P2P evolved initially around file sharing applications. there exists a certain amount of trust and co-operation amongst Grid participants while P2P users have little incentive to act cooperatively. Resource availability tends to be higher and more uniform in the Grid System. Although the number of participating entries in P2P is much higher, the amount of data transferred is much more in Grid based systems (maybe due to better resources) the authors suggest combining the elements of Grid and P2P computing for better computing in the future. From niranjan.sivakumar@gmail.com Thu Apr 20 11:09:09 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-1.9 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from penguin.cs.cornell.edu (penguin.cs.cornell.edu [128.84.96.11]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3KF99216425 for ; Thu, 20 Apr 2006 11:09:09 -0400 (EDT) Received: from xproxy.gmail.com ([66.249.82.205]) by penguin.cs.cornell.edu with Microsoft SMTPSVC(6.0.3790.1830); Thu, 20 Apr 2006 11:08:21 -0400 Received: by xproxy.gmail.com with SMTP id s19so113399wxc for ; Thu, 20 Apr 2006 08:08:21 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:date:from:to:subject:message-id:x-mailer:mime-version:content-type:content-transfer-encoding; b=PWLzJLwJN3cdwA/ngWlCGDzJqsv5Wb8VnzsHxRvR/5f+uxEVMz+mNA0a95Wc513+SSBvqAp+SXBYoMkskS3v4ZGwefxXq6IPiWu8QZCOgJdGC9EgC4oxYmK4pB+/tQSKasbRsfvQ27oBPb/7uHxH/C/Sk7QijZhz7impY+xpx1s= Received: by 10.70.18.13 with SMTP id 13mr78000wxr; Thu, 20 Apr 2006 08:01:46 -0700 (PDT) Received: from localhost ( [69.207.49.126]) by mx.gmail.com with ESMTP id i39sm839315wxd.2006.04.20.08.01.46; Thu, 20 Apr 2006 08:01:46 -0700 (PDT) Date: Thu, 20 Apr 2006 11:01:46 -0400 From: Niranjan Sivakumar To: egs+summary@cs.cornell.edu Subject: PAPER 23 Message-Id: <20060420110146.8ac1a6f6.niranjan.sivakumar@gmail.com> X-Mailer: Sylpheed version 2.2.4 (GTK+ 2.8.13; i686-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-OriginalArrivalTime: 20 Apr 2006 15:08:21.0384 (UTC) FILETIME=[434B3880:01C6648C] Niranjan Sivakumar The Physiology of the Grid On Death, Taxes, and the Convergence of Peer-to-Peer and Grid Computing The Open Grid Services Architecture (OGSA) is designed to provide a method to pool resources and create "virtual organizations" out of heterogeneous platforms. The OGSA model is closely modeled on, and in some ways an extension or evolution of, web services and current uses of the Globus toolkit. OGSA proposes conventions by which services can be described, advertised, and interoperate. The service oriented model is slightly different previous work that tended to focus on physical resources. The example provided in the paper bases service descriptions on WSDL (although it is mentioned that other approaches may also be feasible.) The system allows for different platforms and languages to be used to create the actual services. A number of standard interfaces are defined that may be provided by a grid service. There are provisions to create temporary services and to destroy them. Much as with web services, a client will contact a service provider and obtain information about the services that it provides. Then, the client can authenticate if necessary and create an instance of the service that it requires. The service provider will perform functions that are invoked and respond to the client when necessary. Grid service instances are given a unique Grid service handle (GSH) and this has an associated Grid service reference. While the GSH is constant, the GSR is dynamic and can be changed to reflect upgrades or changes to the service. As with selecting the platform and language to run a service on, OGSA also allows for different network protocols to be used depending on the requirements of the application. P2P and Grid network computing are shown to have many similar properties, but with slightly different focuses. Both P2P networks and Grid computing seek to organize resources across virtual organizations, but P2P network development has focused on failure while Grid development has focused on infrastructural issues. The paper argues that these properties are complementary, and thus there will be an inevitable convergence of the technologies. The authors argue that up to now, the divergence in development has been spurred by different user communities and uses for Grid and P2P networks. Grid computing has largely focused on the scientific community and pooling the resources of relatively well connected, powerful machines for computational purposes. P2P networks have focused on average Internet users and applications such as file sharing, but connecting weaker, less reliably-linked machines. Nonetheless, as these services continue to develop and provide new functionality, their development goals are beginning to converge. OGSA provides a fairly high-level specification of a network and an extremely ambitious goal for it. The biggest problem with OGSA might be that the developers seem to be "over-planning" and in some ways counting their chickens before they hatch. It is not clear if it is really possible to develop a system with the goal of perhaps becoming as popular as the world wide web has become. Furthermore, while they propose some applications for using OGSA, there is no presentation of a real "killer application." The paper on the convergence of P2P and Grid computing offers an interesting observation, but perhaps could benefit for more specific examples of adapting features offered by P2P networks to Grids or vice-versa. The paper stops at a very high level discussing the topic at a fairly general level. From gp72@cornell.edu Thu Apr 20 11:15:20 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-2.1 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from postoffice10.mail.cornell.edu (postoffice10.mail.cornell.edu [132.236.56.14]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3KFFK217840 for ; Thu, 20 Apr 2006 11:15:20 -0400 (EDT) Received: from orpheus3.dataserver.cornell.edu (orpheus3.dataserver.cornell.edu [128.253.161.167]) by postoffice10.mail.cornell.edu (8.12.10/8.12.6) with ESMTP id k3KFFHb8021041; Thu, 20 Apr 2006 11:15:18 -0400 (EDT) Message-ID: <639563208.1145546117592.JavaMail.webber@orpheus3.dataserver.cornell.edu> Date: Thu, 20 Apr 2006 11:15:17 -0400 (EDT) From: Gopal Parameswaran To: egs+summary@cs.cornell.edu Subject: PAPER 23 Cc: gp72@cornell.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: uPortal WEB email client 3.0 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by sundial.cs.cornell.edu id k3KFFK217840 Grid Computing and Peer to Peer Systems Grid Computing and Peer to Peer Systems are two different approaches to highly similar objectives i.e. the pooling and sharing of large sets of distributed resources but differ in their requirements since they focus on two different sets of users. Also Grid computing is primarily focused on distributed computing while Peer to Peer systems currently are restricted to mostly file sharing with focus on anonymities and dealing with failures in the network. Thus a merger of these two areas could deliver better distributed applications that could expand its horizons beyond those of current grid computing resources. In the physiology of the Grid the , the authors discusses the Open Grid Service Architecture that serves as the standard for Grid computing and how Grid computing streamlines the process by using standards and extends the system beyond languages and operating systems by using a web service based architecture. On the other hand peer to peer systems do not currently have any standards developed and hence most applications have their own way of splitting and hashing files and their own network topologies. Thus any user of a peer to peer system if he wants to use two different peer to peer applications need to run both of them simultaneously and cannot share files from one system to another. This limitation in peer to peer system stems from the emerging technology scenario that exists for peer to peer systems as each new implementation of a peer to peer system hold usually little resemblance to a previous system in terms of operation and overlay connections other than supporting the basic functionalities of node join and churn. Grid Computing on the other hand is a commercial solution to enhancing resources for distributed computing. Also since Grid computing uses SOAP which is an XML based protocol over HTTP that provides a means for messaging between a service provider and a service requestor. Thus Grid computing has a more service oriented architecture implemented on a smaller scale where some indexing of ser have a centralized system of indexing service providers whereas peer to peer systems are a totally decentralized system which are scalable across millions of nodes but suffer more from churn related issues. From km266@cornell.edu Thu Apr 20 11:49:15 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-1.7 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from penguin.cs.cornell.edu (penguin.cs.cornell.edu [128.84.96.11]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3KFnE226474 for ; Thu, 20 Apr 2006 11:49:14 -0400 (EDT) Received: from authusersmtp.mail.cornell.edu ([128.253.83.141]) by penguin.cs.cornell.edu with Microsoft SMTPSVC(6.0.3790.1830); Thu, 20 Apr 2006 11:48:42 -0400 Received: from KEVSTOY (cpe-69-207-37-246.twcny.res.rr.com [69.207.37.246]) (authenticated bits=0) by authusersmtp.mail.cornell.edu (8.13.1/8.12.10) with ESMTP id k3KFmcQG028347 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NOT) for ; Thu, 20 Apr 2006 11:48:39 -0400 (EDT) Message-ID: <001f01c66491$eb942d30$f625cf45@KEVSTOY> Reply-To: "Kevin" From: "Kevin" To: Subject: PAPER 23 Date: Thu, 20 Apr 2006 11:48:50 -0400 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2869 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2869 X-OriginalArrivalTime: 20 Apr 2006 15:48:42.0922 (UTC) FILETIME=[E6A498A0:01C66491] The Open Grid Services Architecture (OGSA) if presented in what seems less like a technical paper and more like a call to arms. There are resources being wasted and interoperability between platforms is practically non-existant. Grid responds by setting up a protocol that will allow Virtual Organizations (VO) to share resources with each other (these resources could be memory, processor, whatever). The authors assure quality of service and robustness while trying to build upon currently accepted standards. By allowing all resources to be seen as services, the authors set up Grid to allow for the management of distinct global resources. Death and taxes draw a parallel between grids and p2p. They claim that grids have been focused mainly on infrastructure. Grids, in general, are used on networks where the structure of the network is known or can be administered. Furthermore, the systems are generally up for long periods and do not go down unannounced (unless an error occurs). P2P systems, on the other hand, have been focused on failure. With a high churn rate, P2P systems focus mainly on making sure the system does not go haywire when a quarter of the population randomly disappears. The authors claim that these two systems should and will eventually converge. Grid computing will gain something by churn resistance and P2P system will gain by having a solid infrastructure. From kelvinso@gmail.com Thu Apr 20 13:53:11 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-2.5 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.0 X-Spam-Level: Received: from penguin.cs.cornell.edu (penguin.cs.cornell.edu [128.84.96.11]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3KHrA226029 for ; Thu, 20 Apr 2006 13:53:10 -0400 (EDT) Received: from wproxy.gmail.com ([64.233.184.235]) by penguin.cs.cornell.edu with Microsoft SMTPSVC(6.0.3790.1830); Thu, 20 Apr 2006 13:52:41 -0400 Received: by wproxy.gmail.com with SMTP id i4so438465wra for ; Thu, 20 Apr 2006 10:52:41 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:mime-version:content-type:content-transfer-encoding:content-disposition; b=qf4TaQf4ulnxJlfZPuUmu/Cn+clDg0OWpD/fy3TN45EmKXISwEZSFje8To+9ksf7ISS17uItkNcyOW99nyuf1Icjya0Ftc0zqHqQb9UxcXVzlLX+gXuKkvnEvQ9uPDzBVMLkMu1urvg24AqH5Z0hWXq9daBjuaGrhmpSbUImGo8= Received: by 10.54.108.14 with SMTP id g14mr426480wrc; Thu, 20 Apr 2006 09:03:47 -0700 (PDT) Received: by 10.54.79.14 with HTTP; Thu, 20 Apr 2006 09:03:47 -0700 (PDT) Message-ID: <6e1ca4560604200903v376ee6d4u681e0503184eca3e@mail.gmail.com> Date: Thu, 20 Apr 2006 12:03:47 -0400 From: "Chiu Wah Kelvin So" To: egs+summary@cs.cornell.edu Subject: Paper 23 MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Disposition: inline X-OriginalArrivalTime: 20 Apr 2006 17:52:42.0009 (UTC) FILETIME=[38AF3090:01C664A3] Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by sundial.cs.cornell.edu id k3KHrA226029 The first paper, "On Death, Taxes, and the Convergence of Peer-to-Peer and Grid Computing," compares and contrasts peer-to-peer technology and grid computing. The authors point out differences in the two technologies in their incentive, resources, application, scale, and services that they provided. The target communities are different in the two technologies. In Grid, it is targeted toward scientific computing and business use whereas in P2P it is targeted to diverse and anonymous individual. Therefore, in Grid resource is usually well managed, powerful and available most of the time, and they don't have as much scalability and resilience to failure as P2P. In Grid, the services that they provide are usually persistent and multipurpose whereas in P2P it provides very specific simple services such as file-sharing. Although they are different in many aspects, they both try to solve similar problem which is the sharing of resource within virtual communities. Finally, it concludes that that the two communities are approaching each other over time. The second paper, "The Physiology of the Grid," present an Open Grid Services Architecture that integrate services across distributed, heterogeneous, and dynamic resources from "virtual organization". This paper focuses on three aspects of how a Grid functions, can be implemented and applied. First, it views Grid as an extensible set of Grid services that can be aggregated to achieve the need of various "Virtual Organization." Second, it uses Globus Toolkit to define WSDL interfaces for a Grid services, such as lifetime management, notification, policy management, credential management and virtualization. Finally, they focus on using Grid in commercial setting instead of scientific computing, in which it needs to support integration with existing resources and applications. From lackhand@gmail.com Fri Apr 21 05:49:19 2006 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on sundial.cs.cornell.edu X-Spam-Status: No, score=-1.2 required=5.0 tests=AWL,BAYES_00,HTML_00_10, HTML_MESSAGE autolearn=no version=3.1.0 X-Spam-Level: Received: from iago.cs.cornell.edu (iago.cs.cornell.edu [128.84.96.10]) by sundial.cs.cornell.edu (8.11.7-20031020/8.11.7/M-3.25) with ESMTP id k3L9nI201365 for ; Fri, 21 Apr 2006 05:49:19 -0400 (EDT) Received: from pproxy.gmail.com ([64.233.166.183]) by iago.cs.cornell.edu with Microsoft SMTPSVC(6.0.3790.1830); Fri, 21 Apr 2006 05:49:12 -0400 Received: by pproxy.gmail.com with SMTP id c30so420284pyc for ; Fri, 21 Apr 2006 02:49:12 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:mime-version:content-type; b=fz2/7QT6tfzdTFtrQdZqPTNRbd7f18G5DDjuhxxO3DKYNZTkiG1KJiaMFB7p8bvvjqkMtC1ck7oYYOzwt8pQplqNxmHg6RbPlAZE+li1x1MlncqmJKBI6IPNvhtwJHHZaxaBLvbEhS7eXF1mJmTuQKeXkrc0g53bNt4TNzPrSxE= Received: by 10.35.36.13 with SMTP id o13mr910982pyj; Thu, 20 Apr 2006 21:21:33 -0700 (PDT) Received: by 10.35.125.16 with HTTP; Thu, 20 Apr 2006 21:21:33 -0700 (PDT) Message-ID: <9aa7a97d0604202121s8345357r1b40634efb88314a@mail.gmail.com> Date: Fri, 21 Apr 2006 00:21:33 -0400 From: "Andrew Cunningham" To: egs+summary@cs.cornell.edu Subject: PAPER 23 MIME-Version: 1.0 X-Security: message sanitized on sundial.cs.cornell.edu See http://www.impsec.org/email-tools/sanitizer-intro.html for details. $Revision: 1.148 $Date: 2004-12-19 11:59:17-08 X-Security: The postmaster has not enabled quarantine of poisoned messages. Content-Type: multipart/alternative; boundary="----=_Part_25827_7858477.1145593293370" X-OriginalArrivalTime: 21 Apr 2006 09:49:12.0996 (UTC) FILETIME=[D860D640:01C66528] ------=_Part_25827_7858477.1145593293370 Content-Type: text/plain; charset=ISO-8859-1 Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Forgive the lateness -- I returned home to find this in my drafts folder! Andrew Cunningham arc39 The Physiology of the Grid Ian Foster, Carl Kesselman, Jeffrey M. Nick, Steven Tuecke Similar to the Death and Taxes paper, rather than exploring the differences and similarities between peer to peer and grid systems, this paper focuses on the technical specifications of grid systems and applications. It presents the Globus toolkit and web services, along with a bevy of buzzwords, which are real and implemented technologies that work together to permit interoperability of disparate services. The ultimate point of this work is to permit virtual organizations to cohere; in other words, to allow individual entities to negotiate efficiently how to communicate with each other. The Grid service is represented as a sort of server that operates in many ways as a kernel, providing service handles to client processes via invocations to default portions of the Grid service that are themselves handles. On Death, Taxes, and the Convergence of Peer-to-Peer and Grid Computing Ian Foster, Adriana Iamnitchi Grid computing is a different approach to resource sharing across networks, with different foci and assumptions that determine their target audiences. The differrence between this and P2P is the degree of centralization and location of control, in that peer to peer systems tend t= o be allergic to centralization, where Grid Computing tries to use its resources as efficiently as possible, while accepting centralized control. Grid applications focus on large organizations supplying resources and make= s more assumptions about security, trust, and up-time. Grid computing also is concerned with providing standardized interfaces to allow generalized resource sharing, where Peer to Peer systems focus on vertically integrated solutions to specific resource sharing issues. Breaking standard form, this final paragraph comments on both papers. They are somewhat content free, providing descriptions of real systems that are separate architectures in and of themselves, but which do not necessarily add up to the suggested technology. Grid permits the sharing of resources, but there is so much noise and so much tinsel that it is hard to see exactly where this occurs; the heavy lifting seems to be done by other, less exalted technologies. ------=_Part_25827_7858477.1145593293370 Content-Type: text/html; charset=ISO-8859-1 Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Forgive the lateness -- I returned home to find this in my drafts folder!
  
Andrew Cunningham
arc39
    The Physiology of the Grid
    Ian Foster, Carl Kesselman, Jeffrey M. Nick, Steven Tuec= ke
    Similar to the Death and Taxes paper, rather than exploring the differences and similarities between peer to peer and grid systems, this paper focuses on the technical specifications of grid systems and applications. It presents the Globus toolkit and web services, along with a bevy of buzzwords, which are real and implemented technologies that work together to permit interoperability of disparate services. The ultimate point of this work is to permit virtual organizations to cohere; in other words, to allow individual entities to negotiate efficiently how to communicate with each other. The Grid service is represented as a sort of server that operates in many ways as a kernel, providing service handles to client processes via invocations to default portions of the Grid service that are themselves handles.
    
    On Death, Taxes, and the Convergence of Peer-to-Peer and= Grid Computing
    Ian Foster, Adriana Iamnitchi
    Grid computing is a different approach to resource sharing across networks, with different foci and assumptions that determine their target audiences. The differrence between this and P2P is the degree of centralization and location of control, in that peer to peer systems tend to be allergic to centralization, where Grid Computing tries to use its resources as efficiently as possible, while accepting centralized control. Grid applications focus on large organizations supplying resources and makes more assumptions about security, trust, and up-time. Grid computing also is concerned with providing standardized interfaces to allow generalized resource sharing, where Peer to Peer systems focus on vertically integrated solutions to specific resource sharing issues.
    
    Breaking standard form, this final paragraph comments on both papers. They are somewhat content free, providing descriptions of real systems that are separate architectures in and of themselves, but which do not necessarily add up to the suggested technology. Grid permits the sharing of resources, but there is so much noise and so much tinsel that it is hard to see exactly where this occurs; the heavy lifting seems to be done by other, less exalted technologies. ------=_Part_25827_7858477.1145593293370--