Hartman, J., et al., "Liquid Software: A New Paradigm for Networked Systems," June 1996, Department of Computer Science, University of Arizona.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....disks, service migration, multiple views, virtual block addresses. 1INTRODUCTION M ANY researchers have suggested migrating services to storage devices to improve overall system performance. A number of studies have demonstrated the benefits of migrating services to different levels in a system [1], 2] 3] 4] These studies have motivated the utility of devices with more intelligence and function. Direct network attachment of disks is being proposed to enable data transfers from devices directly to clients rather than through servers [5] 6] 7] 8] Earlier work on active disks [9] ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting, "Liquid Software: A New Paradigm for Networked Systems," Technical Report TR96-11, Dept. of Computer Science, Univ. of Arizona, Tucson, 1996.

....fine grained code mobility between active switches. These projects build on existing work such as the MIT Active Node Transfer System (ANTS) 22] and University of Pennsylvania s SwitchWare architecture [2] In addition, there is ongoing work to develop general OS support for liquid software [9, 19]. One application of Xenoservers is to host a test deployment of these existing mobile agent platforms at strategic places within the Internet, providing a common substrate for the authentication and resource management that they require for public use. Summary. Existing work provides piece wise ....

HARTMAN,J.,MANBER,U.,PETERSON,L.,AND PROEBSTING, T. Liquid software: A new paradigm for networked systems. Tech. Rep. TR96-11, Department of Computer Science, University of Arizona, 1996.

....parallelism on I O performance. These systems still treat disks as merely as storage. Our work will extend the function of any disk on the network to encompass local processing at the disk. Active Networks [38, 30] envisions code migration to the network to improve QOS. These projects [9, 17] deal with execution of user code (carried within network packets) in the network. The migratable code is much more application specific in smart disk systems as opposed to dealing with a number of predefined QOS attributes of the network. 8 Conclusions Putting excessive computational power to ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting. Liquid software: A new paradigm for networked systems. Technical Report 96-11, Department of Computer Science, University of Arizona, 1996.

....serverlevel parallelism on I O performance. These systems still treat disks as merely as storage. Our work will extend the function of any disk on the network to encompass local processing at the disk. Active Networks [39, 30] envisions code migration to the network to improve QOS. These projects [10, 18] deal with execution of user code (carried within network packets) in the network. The migratable code is much more application specific in smart disk systems as opposed to dealing with a number of predefined QOS attributes of the network. 7 Conclusions Putting excessive computational power to ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting. Liquid software: A new paradigm for networked systems. Technical Report 96-11, Department of Computer Science, University of Arizona, 1996.

....parallelism on I O performance. These systems still treat disks as merely as storage. Our work will extend the function of any disk on the network to encompass local processing at the disk. Active Networks [38, 30] envisions code migration to the network to improve QOS. These projects [9, 17] deal with execution of user code (carried within network packets) in the network. The migratable code is much more application specific in smart disk systems as opposed to dealing with a number of predefined QOS attributes of the network. 8. CONCLUSIONS Putting excessive computational power to ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting. Liquid software: A new paradigm for networked systems. Technical Report 96-11, Department of Computer Science, University of Arizona, 1996.

....of active network devices, there is no agreement on which primitives must be provided by them. Moreover, researchers presently interpret the idea of active network at least with two different nuances. The first, broader, interpretation of the term, embodied for example in the work described in [47, 128], is that the distinction between the nodes inside and those at the edge of a network becomes blurred in terms of their functional characteristics. In this setting, network devices, e.g. routers, can execute directly mobile code of a distributed application, providing it with the capability to ....

Hartman, J., Manber, U., Peterson, L., and Proebsting, T. Liquid Software: A New Paradigm for Networked Systems. Tech. Rep. 96-11, Dept. of Computer Science, Univ. of Arizona, Tucson, AZ, June 1996.

....a single chip has also been applied to memories. For example, Berkeley s IRAM (intelligent RAM) project [36] implements such an integration which reduces memory latency and increases memory bandwidth. Active Networks [28, 23] envisions code migration to the network to improve QOS. These projects [7, 14] deal with execution of user code (carried within network packets) in the network. The migratable code is much more application specific in smart disk systems as opposed to dealing with a number of predefined QOS attributes of the network. 7 Conclusions A growing trend in computer architecture ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting. Liquid software: A new paradigm for networked systems. Technical Report 96-11, Department of Computer Science, University of Arizona, 1996.

....have also developed a new language called PLAN [20] which is written specifically for the Active Networking platforms. The idea behind developing Netscript project [21] at Columbia University is similar. Netscript provides a medium for creating programmable networks. The Liquid Software project [22][23] at the University of Arizona aims to develop technologies which allow use of mobile code inside the network. The Active Networking project at Georgia Tech [24] embeds vendor defined functions at network nodes. These functions are available to the user for performing user specific services. ....

J. Hartman et al. Liquid Software: A New Paradigm for Networked Systems. Technical Report TR96-11, Dept. of Computer Science, Univ. of Arizona, 1996.

....approach. 1 Introduction Many researchers have suggested various enhancements to storage devices to improve the performance, function and characteristics by migrating services to disks. A number of studies have 1 demonstrated the bene ts of migrating services to di erent levels in the system [1, 2, 3, 4]. These studies have motivated the utility of devices with more intelligence and function. Device level enhancements such as compression, encryption, and log based le systems [5] are being proposed. Direct network attachment of disks is being proposed to enable data transfers from the device ....

John Hartman, Udi Manber, Larry Peterson, and Todd Proebsting. Liquid software: A new paradigm for networked systems. Technical report, Department of Computer Science, The University of Arizona, 1996. 27

....Active Networks themes to the accounting coprocessor. AIACE, however, is limited to the management plane, it does not interfere with the end to end path, and thus it may find paths of lesser resistance towards adoption. Likewise, AIACE can be regarded as an original application of Liquid Software [8] to a path that is not the end to end path, but is rather the IP accounting and management path. In [13] the accounting billing infrastructure meets Active Networks with a radical new approach: the accounting and billing functions are delegated to customer s systems. AIACE shares the same ....

J. Hartman, et al. "Liquid Software: A new Paradigm for Networked Systems," Technical Report 96-11, Dept. of Computer Science, The University of Arizona.

....Researchers there have also developed a new language called PLAN [19] written specifically for active networking platforms. The idea behind the Netscript project [20] at Columbia University is to provide a scripting language for creating data flow paths in the network. The Liquid Software project [21] at the University of Arizona investigates technologies that allow use of mobile code inside the network. The active networking project at Georgia Tech [22] embeds vendor defined functions at network nodes. These functions are available to the user for performing user specific services. Data ....

J. Hartman, U. Manber, L. Peterson and T. Proebsting, Liquid Software: A New Paradigm for Networked Systems, Technical Report TR96-11, Dept. of Computer Science, Univ. of Arizona, 1996.

....Java is used in all commercial systems and in several research systems. Due to the recognition that agents must execute at near native speed to be competitive with traditional techniques in certain applications, however, several researchers are experimenting with onthe y compilation [LSW95, HMPP96] The agent initially is compiled into bytecodes, but compiled into native code on each machine that it visits, either as soon as it arrives or while it is executing. The most recent Java virtual machines use on the y compilation, and the Java only mobile agent systems, which are not tied to a ....

John Hartman, Udi Manber, Larry Peterson, and Todd Proebsting. Liquid software: A new paradigm for networked systems. Technical Report TR96-11, Department of Computer Science, University of Arizonia, 1996.

....know what methods this object exports Bob s phone cannot store interfaces for every device in his home, so it must request the correct interface whenever it needs to execute a remote method. How is this accomplished Several emerging solutions to this problem, including Jini, Liquid Software [13], and the Active Networks Toolkit [40] are based on the Java language. In these models, bytecode is downloaded and executed on the client. Using the Java RMI, clients can then use the services of other devices. Other solutions are based on CORBA [12] and Microsoft s COM [19] Both allow clients to ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting. Liquid software: A new paradigm for networked systems. Technical report, University of Arizona, Tuscon, AZ, 1996.

....that, packets have the option to specify whether to use the default routing protocol or a customised routing protocol. Issues a ecting the performance of their architecture include: kernel crossings, use of threads, copying and garbage collection cost. The paradigm envisioned by Hartman et al. [158] is similar to ANs. In [158] they attempt to apply mobile code technologies to all layers of the communication stack. Instead of a speci c code, executing on a speci c machine, the code ows from one machine to another, thus the term liquid software. The aim here is to transport and transform ....

....to specify whether to use the default routing protocol or a customised routing protocol. Issues a ecting the performance of their architecture include: kernel crossings, use of threads, copying and garbage collection cost. The paradigm envisioned by Hartman et al. 158] is similar to ANs. In [158], they attempt to apply mobile code technologies to all layers of the communication stack. Instead of a speci c code, executing on a speci c machine, the code ows from one machine to another, thus the term liquid software. The aim here is to transport and transform data eciently. One main ....

J. H. Hartman, U. Manber, L. L. Peterson, and T. A. Proebsting, \Liquid software: A new paradigm for networked systems," tech. rep., Department of Computer Science, University of Arizona, 1996. Technical Report 96-11.

....their particular requirements. These studies have motivated the need for devices with more intelligence and function. It is likely that di erent enhancements may be proposed in the future. A number of other studies have motivated the bene ts of migrating services to di erent levels in the system [10, 11, 12, 13]. Stackable le system model [14, 15, 16] has been suggested to provide services at le system (FS) level. However, in these systems, services can only be supported at the le system level. Some approaches associate services with the process of name resolution [11, 17] Other approaches have ....

....speci ed arguments. The idea of extending le system services through name space is further generalized in Active Names [11] A number of approaches for supporting computation migration are proposed in Active Names [11] Active networks [12] Sprite [42] Active Messages [43] Liquid software [10, 18], Derived Virtual Devices [44, 13] Most of these approaches allow service migration only above the le system. 5 Conclusions We have proposed to use multiple views of storage to build a exible storage system. The multiview storage system, MVSS, provides multiple views of an underlying le, ....

John Hartman, Udi Manber, Larry Peterson, and Todd Proebsting. Liquid software: A new paradigm for networked systems. 1996.

....such as Jini [Sun99] for managing and configuring devices builds on this same advantage to address a particular part of the problem, mediating the interaction among heterogeneous devices. There are a number of additional domains where a general purpose mobile code system would be applicable [Hartman96]. 2.5.2 Virtual Machines The use of a virtual machine provides two complimentary benefits, the first is the ability to use the same program on a variety of underlying machine and processor architectures, the second is the greater degree of controlled provided in a virtual machine, when the code ....

Hartman, J., Manber, U., Peterson, L. and Proebsting, T. "Liquid Software: A New Paradigm for Networked Systems" Technical Report 96-11, University of Arizona, 1996.

....software developers to interact through a set of well defined protocols. Active networking is motivated by the notion that the improvement and evolution of current networking software is greatly hindered by slow and expensive standardization processes. Several active networking research projects [1, 4, 5, 12 14, 11] try to accommodate changes to network software by facilitating the safe and efficient dynamic reconfiguration of the network. Adaptive networks may be seen as the composition of the two main orthogonal approaches to active network design discussed in [13] In the discrete approach ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting. Liquid software: A new paradigm for networked systems. Technical Report 96-11, University of Arizona, 1996.

....is complementary to several other active network efforts. The use of general purpose Java bytecodes and VM has allowed us to evolve our architecture quickly, but at the cost of less control over resource usage and lower absolute performance. Research at the University of Arizona on Liquid Software [8] and Scout [11] enable a finer granularity of local resource management as well as competitive performance through the construction of a specialized node operating system. Research at the University of Pennsylvania on PLAN and BBN on Sprocket enables stronger resource management and security ....

J. Hartman et al. Liquid Software: A New Paradigm for Networked Systems. Technical Report TR96-11, Dept. of Computer Science, Univ. of Arizona, 1996.

.... ,C B ,DA REV A,DA ,C A B,DB ,C B A,DA ,C A B,DB ,C B ,DA ,C A COD A,DA ,C A B,DB ,C B A,DA ,C A ,C B ,DB B,DB ,C B MA A,DA ,C A DB ,C B DA ,C A A,DB ,C B ,DA ,C A Table 1: Mobility and remote execution paradigms [Whi96a] Voyager[Obj96] and the Liquid Software project at University of Arizona[HMPP96] In the former category are languages whose semantics were designed to support distributed scope and access, provide a shared memory abstraction, and extend ordinary language operations to manage replicated objects and data. The latter category comprise agent languages and mobile code frameworks ....

J. Hartman, U. Manber, L. Peterson, and T. Proebsting. Liquid software: A new paradigm for networked systems. Technical Report Technical Report 96-11, Univ. of Arizona, June 1996.

No context found.

Hartman, J., et al., "Liquid Software: A New Paradigm for Networked Systems," June 1996, Department of Computer Science, University of Arizona.

No context found.

J. Hartman, U. Manber, L. L. Peterson and T. Proebsting, "Liquid Software: A New Paradigm for Networked Systems, Technical Report", Dept of Computer Science, The University of Arizona 1996.

No context found.

Hartman, J., Manber, U., Peterson, L., and Proebsting, T., "Liquid software: a new paradigm for networked systems," Tech. Rep. 96-11, Department of Computer Science, University of Arizona, Tucson, AZ, June 1996.

No context found.

J. Hartman, U. Manber, L. Peterson, T. Proebsting. "Liquid Software: A New Paradigm for Networked Systems." Technical Report 96-11. Dept. of Computer Science, The University of Arizona.

No context found.

Hartman, J., et al., "Liquid Software: A New Paradigm for Networked Systems", Technical Report 96-11, Dept. of Computer Science, Univ. of Arizona, 1996.

No context found.

John Hartman et al., Liquid Software: A new Paradigm for Networked Systems, Technical Report TR96-11, Department of Computer Science, University of Arizona, 1996.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC