K. Birman and K. Marzullo. ISIS and the META Project. Sun Technology, vol. 2, no. 3 (Summer 1989), pp. 90-104.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....C compiler for the stubs, perhaps others for the node computations. For each node, the two components are linked together to produce an executable module. The compiler must also address the two aspects of heterogeneity data representation and instruction sets. Paralex uses the ISIS toolkit [9] as the infrastructure to realize a universal data representation. All data that is passed from one node to another during the computation are encapsulated as ISIS messages. Paralex automatically generates all necessary code for encodingdecoding basic data types (integer, real, character) and ....

K. Birman and K. Marzullo. ISIS and the META Project. Sun Technology, vol. 2, no. 3 (Summer 1989), pp. 90-104.

....Fellowship. Intel Corporation. k IBM Canada Laboratory. These ideas are not new; parallel computing has long been an active area of research. The fact that networks of computers are commonly being used in this fashion is new. Software tools like PVM [1, 15] P4 [5] Linda [7] Isis [2], Express [14] and MPI [16] allow a programmer to treat a heterogeneous network of computers as a parallel machine. These tools are useful, but for efficient and practical use, load balancing and fault tolerance mechanisms must be developed that will work well in a heterogeneous multi user ....

K. Birman and K. Marzullo. Isis and the META project. Sun Technology, pages 90--104, Summer 1989.

....case is that of managementof# groups of objects. This has been recognised in projects such as Domino which uses the notionof# management domains for this purpose. However, the concept of groups isimportant# in distributed systems irrespective of it srole# in management. Systems such as ISIS [3] have demonstrated the effectiveness of process groups as a programming abstraction, and object groups area# key part of our system. Hence, rather thanadopt# a specialised means of groupingfor# management purposes we intend to utilise the powerof# object groups. As a precursor to this work wehave# ....

Birman, K., and Marzullo, K., "ISIS and the META Project." SUN Technology SUN Technology No.: Summer 1989,

....an even more powerful computing resource than any single machine. These ideas are not new; parallel computing has long been an active area of research. The fact that networks of computers are commonly being used in this fashion is new. Software tools like PVM [1, 7] P4 [11] Linda [4] Isis [2], and Express [6] allow a programmer to treat a heterogeneous network of computers as a parallel machine. These tools allow the programmer to partition a program into pieces which may then execute in parallel, occasionally synchronizing and exchanging data. The tools generally support conversion ....

Kenneth Birman and Keith Marzullo. Isis and the META project. Sun Technology, pages 90--104, Summer 1989.

....only wake the client when information on a particular address was received. Currently none of the architecture has been implemented, although changes to the world wide web protocol (HTTP) have been proposed. 3.6. 3 MobileChannel MobileChannel [Cho,94] is a tool constructed on top of the ISIS [Birman,89] distributed systems platform which aims to support host migration through group communication. ISIS provides two fundamental functions: group management and group communication. Two communications primitives are supported for group communication: CBCAST and ABCAST. CBCAST implements reliable ....

Birman, K., and K. Marzullo. "ISIS and the META Project." Sun Technology, Vol. 2 No. 3. September 1989.

....In contrast, a NOW is a distributed, often heterogeneous resource that is highly shared. Processors can be of different architectures, and are usually running a time sharing operating system. Programs for NOWs are generally written using some NOW programming environment like PVM [16] or Isis [4] that provides process control, message passing, etc. These programming envirnoments allow individual processors to enter or leave the NOW dynamically due to availability, load, or failure and thus present a far more flexible failure model than supercomputers. In such systems, consistent ....

....the processors in G j , for 0 j m. This is basically a 1 dimensional parity scheme, which can tolerate up to m simultaneous processor failures, as long as each failure occurs in a different group [18] The extreme we have presented is m = 1. At the other extreme are systems like Isis [4] or Targon [7] where m = N , and every processor has a backup processor to which it sends checkpoints. As m grows, the overhead of checkpointing and recovery will decrease because there is less contention for the parity processors. Network of Sparc 2 s: k Running With Checkpointing Recovery Time ....

K.P. Birman and K. Marzullo. ISIS and the meta project. Sun Technology, Summer 1989.

....causal broadcast, so that all the copies of a process can determine their stable input messages [Schn90] Some distributed systems exist which support programming with process replications. Probably, the best known of them is PARALEX [Baba92] which has been built on top of the ISIS toolkit [Birm89], and uses its coordinator cohort library to implement passive replication. However, PARALEX cannot be used for implementing TW distributed simulation, because its model only allows the definition of processes without persistent internal status and a statically defined communication acyclic graph. ....

K. Birman, K. Marzullo, "ISIS and the META Project", SUN Technology, 2(3), pp.90-104, 1989.

....the processors in G j , for 0 j m. This is basically a 1 dimensional parity scheme, which can tolerate up to m simultaneous processor failures, as long as each failure occurs in a different group [21] The extreme we have presented is m = 1. At the other extreme are systems like Isis [5] or Targon [7] where m = n, and every processor has a backup processor to which it sends checkpoints. As m grows, the overhead of checkpointing and recovery will decrease because there is less contention for the parity processors. To tolerate any combination of m processor failures, m parity ....

K.P. Birman and K. Marzullo. ISIS and the meta project. Sun Technology, Summer 1989.

....all UBLCS 92 1 3 Fault Tolerance of the stub code as ordinary C. As the next step, the C compiler is invoked to turn each node into an executable module. The compiler must also address the two aspects of heterogeneity: Data representation and instruction sets. Paralex uses the ISIS toolkit [6,7] as the infrastructure to realize a universal data representation. All data that is passed from one node to another during the computation are encapsulated as ISIS messages. Heterogeneity with respect to instruction sets is handled in a brute force manner by invoking remote compilations on the ....

K. Birman and K. Marzullo. ISIS and the META Project. Sun Technology, vol. 2, no. 3 (Summer 1989), pp. 90--104.

....an even more powerful computing resource than any single machine. These ideas are not new; parallel computing has long been an active area of research. The fact that networks of computers are commonly being used in this fashion is new. Software tools like PVM [1, 13, 14] P4 [5] Linda [6] Isis [2], Express [12] and MPI [16] allow a programmer to treat a heterogeneous network of computers as a parallel machine. These tools allow the programmer to partition a program into pieces which may then execute in parallel, occasionally synchronizing and exchanging data. Heterogeneity is supported ....

Kenneth Birman and Keith Marzullo. Isis and the META project. Sun Technology, pages 90--104, Summer 1989.

....Paralex generates all of the stub code as ordinary C. As the next step, the C compiler is invoked to turn each node into an executable module. The compiler must address two aspects of heterogeneity in a distributed system: data representation and instruction sets. Paralex uses the ISIS toolkit [10,11] as the infrastructure to realize a universal data representation. All data that is passed from one node to another during the computation are encapsulated as ISIS messages. Heterogeneity with respect to instruction sets is handled by invoking remote compilations on the machines of interest and ....

K. Birman and K. Marzullo. ISIS and the META Project. Sun Technology, vol. 2, no. 3 (Summer 1989), pp. 90--104.

....program. Once the collection of processors comprising the machine was selected, processors could not be added or deleted, and individual processor failures rendered the entire system useless, as in a supercomputer. In recent NOW programming environments, like PVM version 3 [GBD 94] and Isis [BM89] the virtual machine can tolerate change. Processors can be added to or removed from the collection and failures are tolerated. In such systems, consistent checkpointing is overkill. If one processor fails, the whole collection of processors must restart themselves from disk. Moreover, if the ....

....processors in G j , for 0 j m. This is basically a 1 dimensional parity scheme, which can tolerate up to m simultaneous processor failures, as long as each failure occurs in a different group [GHK 89] The extreme we have presented is m = 1. At the other extreme are systems like Isis [BJ89, BM89] or Targon [BBG 89] where m = N , and every processor has a backup processor to which it sends checkpoints. As m grows, the overhead of checkpointing and recovery will decrease, since the recovery group size is smaller, which necessitates less information to be combined at each parity ....

K. P. Birman and K. Marzullo. ISIS and the meta project. Sun Technology, Summer 1989.

....systems as well. Express 3 [FKB91] is a collection of tools for programming distributed memory multiprocessors, both true multiprocessors and UNIX network clusters. It includes a message passing interface and tools for automatic parallelization of serial code, debugging and visualization. ISIS [BM89] is a message passing programming system for workstation clusters. The system uses a concept called virtual synchrony and knowledge of the level of synchronization necessary between application components to efficiently maintain distributed data structures. Isis applications can be programmed in ....

Kenneth Birman and Keith Marzullo. ISIS and the meta project. SunTechnology, Summer 1989.

....communication, concurrency control, atomicity control, and crash recovery. The services that are present in current operating systems are general purpose and do not satisfy the demands of distributed 1 Unix is a trademark of AT T Bell Laboratories. transaction processing algorithms [2, 3, 5]. For instance, locking facilities and buffer management are generally implemented by database systems because the services provided in operating systems are inadequate. This paper is organized as follows. Section 2 discusses design, implementation, and performance issues of Push. Section 3 ....

Kenneth Birman and Keith Marzullo. ISIS and the META project. Sun Technology, pages 90--104, July 1989.

....a number of simple applications and a platform browser which allows direct user access are currently being developed beyond an existing prototype. The prototype contains a fully working event model and a set of skeleton support services, coupled with a number of initial browser interfaces. ISIS (Birman and Marzullo, 1989) was chosen as a suitable toolkit to support the development in a distributed environment because of its use of process groups and associated multicast facilities. We want to compare and contrast ISIS as an underlying distributed support system with others, such as the ANSA TESTBENCH (ANSA, 89) ....

Birman, K., Marzullo, K. (1989): "ISIS and the META project." Sun Technology , Summer,1989, pp.90-104.

....a suite of user interface primitives (shown in Table 2 in the Appendix) and supporting software that together enable concurrent computing on loosely coupled networks of processing elements. Several design features distinguish PVM from other similar systems such as Cosmic [7] Marionette [4] ISIS [22], and Dpup [5] Among these are the combination of heterogeneity, scalability, multilanguage support, provisions for fault tolerance, the use of multiprocessors and scalar machines, an interactive graphical front end, and support for profiling, tracing, and visual analysis. 2.1. PVM Architecture ....

K. Birman and K. Marzullo, "ISIS and the META project", Sun Technology Summer 1989, pp. 90-104.

....by the Applied Mathematical Sciences program, Office of Basic Energy Sciences, U. S. Department of Energy, under Grant No. DE FG05 91ER25105, and Contract DE AC05 84OR21400 with Martin Marietta Energy Systems, Inc. support tools, and straightforward interfaces. Examples of such systems are Isis [1], Linda [2] Express [3] and PVM [7] It should be noted that these systems, referred to hereafter as environments , are distinct from distributed operating systems, which are native kernel level implementations. Wide ranging experiences (in terms of quantity and diversity of use) during the ....

K. Birman and K. Marzullo, "ISIS and the META project" Sun Technology , pp. 90-94, Summer 1989.

....of Paralex programs. HeNCE graphs are richer than those of Paralex. HeNCE provides dynamically spawned subgraphs, pipelining, loops and conditionals. Pipelining is provided in Paralex but only for the entire program graph. There are several platforms on which HeNCE could have been built. Isis [6] is a parallel programming toolkit for fault tolerant parallel computing over a network of heterogeneous machines. Isis is a large system. It requires significant system resources and a system administrator to properly install. The Cosmic environment [11] is a publicly available parallel ....

Kenneth Birnam and Keith Marzullo. Isis and the META project. Sun Technology, pages 90-- 104, Summer 1989.

....the hosts on a network, and collectively manages its resources. These systems mandate that the distributed kernel must execute on all the system nodes a limiting factor for many applications and environments. Distributed Programming Environments ################################# such as ISIS [3], Marionette [4] PVM [5] and Linda [6] consist of a software layer that executes above the individual operating systems of autonomous machines, and provides various distributed facilities. While these systems are more flexible than distributed operating systems, they too have disadvantages, ....

....of the research efforts that are related to the DCL project are integral to distributed programming systems and environments, and are therefore not directly comparable. Nevertheless, many systems do support subsets (sometimes intersecting) of the facilities provided by DCL. One example is ISIS[3], a distributed environment centered around replication and consistency. ISIS primarily supports the client server paradigm, and uses atomic broadcast as the primary communications mechanism. The Chare kernel [7] is aimed at providing a unified paradigm for both shared and distributed memory ....

K. Birman and K. Marzullo, ISIS and the META project, Sun Technology, 90-104, Summer 1989.

....HeNCE might be expanded to be compatible. Paralex [1] is a system similar to HeNCE, but with further restrictions. Its ability to represent pipelining is constrained to a single program graph, and each procedure is allowed only a single output. Further, it is targeted to, and relies upon, the Isis [7] parallel programming toolkit, whereas Zoom is independent of a particular programming environment. 1.2 A Prototype Heterogeneous Application To motivate key aspects of heterogeneous applications, we describe an application based loosely on the Calcrust application. Calcrust [5] developed at ....

Birman, K., and Marzullo, K. Isis and the meta project. Sun Technology 2, 3 (Summer 1989), 90--104.

....phase. Currently, Paralex generates all of the stub code as ordinary C. As the next step, the C compiler is invoked to turn each node into an executable module. The compiler must also address the two aspects of heterogeneity: Data representation and instruction sets. Paralex uses the ISIS toolkit [9,8] as the infrastructure to realize a universal data representation. All data that is passed from one node to another during the computation are encapsulated as ISIS messages. Heterogeneity with respect to instruction sets is handled in a brute force manner by invoking remote compilations on the ....

K. Birman and K. Marzullo. ISIS and the META Project. Sun Technology, vol. 2, no. 3 (Summer 1989), pp. 90--104.

....studied the following areas; developing software tools, increasing communication performance, and studying feasibility of distributed computing on a SNOW. Several software tools have been developed to support SNOW environments; PVM [Sun90, Gei94] P4 [Boy87] Express [FKB91] Linda [CG89] Isis [BM89], and MPI [GLS94] These tools allow programmers to treat a heterogeneous collection of processors in a SNOW as one large virtual machine. The virtual machine can be composed of a set of identical workstations. But since networked computers can have several other users on them running a variety of ....

Kenneth Birman and Keith Marzullo. "Isis and the META project." Sun Technology, pp. 90--104, Summer 1989.

No context found.

Kenneth Birman and Keith Marzullo. Isis and the META project. Sun Technology, pages 90-- 104, Summer 1989.

No context found.

Birman, K., K. Marzullo "ISIS and the META project", SUN Technology, Summer 1989.

No context found.

Birman, K., K. Marzullo "ISIS and the META project", SUN Technology, Summer 1989.

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