N. Carriero, D. Gelernter. "The S/Net's Linda Kernel". Proceedings Symp. Operating Systems Principles, Dec. 85.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the network. Thus, when a process wants a tuple, it only needs to search in its node, because tuple space is completelly replicated in each node. But the attempt to delete it (by means of an in operation) implies the activation of a global protocol, because the tuple must be deleted in all nodes [8]. This option is viable with good performance only if we have great memory capacity and fast communication medium, mainly with reliable broadcast. 4.1.2 N 1 distribution The inverse choice is also possible. The tuple space can be partitioned into N small parts, each one assigned to one of the ....

N. Carriero, D. Gelernter. "The S/Net's Linda Kernel". Proceedings Symp. Operating Systems Principles, Dec. 85.

....processes can freely move 3 from machine to machine (to balance load or to evict processes when a machine s owner returns) and process signals can be propagated to migrated processes. A few examples of completely distributed operating systems are Accent [10] Amoeba [3] Charlotte [11] Linda [12], Sprite [2] and the V System [13] GLUnix [14] is a compromise between a distributed operating system and a process distribution facility. Written as a layer of software that runs on top of an existing operating system, GLUnix can make a network of computers look like a global operating ....

N. Carriero, D. Gelernter, "The S/Net's Linda Kernel," ACM Transactions on Computer Systems, 4(2):110--129, May 1986.

....of technologies such as JavaSpaces, and merits investigation of their use for high performance computation. 1. 1 Related Work Since their introduction by Gelernter et al. in the 1980 s, tuplespaces have been thoroughly described in the literature and exist in a wide variety of implementations [5, 6, 7, 8]. We present a brief review here. The distinguishing feature of tuplespaces is generative communication, whereby processes do not communicate directly, but rather by adding and removing tuples (ordered collections of data) to and from a space, process independent storage with shared memory like ....

N. Carriero, D. Gelernter, "The S/Net's Linda Kernel", ACM Trans. on Computer Systems, Vol. 4, No. 2, May 1986

....debugger for C programs. Status: Several prototypes of Meglos systems have been built and are being used to support multiprocessor robotics applications in the AT T Robotics Research Department. Contact: Robert D. Gaglianello, AT T Laboratories, Holmdel, NJ 07733. References: 237] [238], 239] 240] 241] 35 2.35 MOS Main Goal The Multicomputer Operating System (MOS) is a distributed UNIX system. It is a general purpose time sharing operating system which makes a cluster of loosely connected independent homogeneous computers behave as a single machine UNIX system. The main ....

....a parallel programming language which supports shared memory like distributed data structures. S Net is a multicomputer based on a fast word parallel bus interconnect. A special communication kernel is implemented 60 for this machine that supports Linda primitives (cf. section 2. 34) Reference: [238] Sesame Sesame is a file system, developed at Carnegie Mellon University as part of the Spice project. It is a successor to CMCFS (cf. this section) Its main objectives are naming, authentication, authorization and data storage as well as retrieval in a network of personal computers. ....

N. Carriero and D. Gelernter, "The S/Net's Linda Kernel", ACM Transactions on Computer Systems, 4(2):110--129, May 1986.

....virtual associative memory, shared by all processes joined in the computation. Linda was implemented in a wide range of platforms, with shared memory architectures, like the Encore, Sequent and Alliant machines as well as with distributed memory, like the Intel iPSC hypercube [38] Bell Labs S Net [8] and transputers[5] 8] There are also implementations in more loosely coupled systems, e.g. VAX VMS Workstation clusters [31] PC Networks [12] and Sun Workstations [24] In spite of its wide acceptance, there have been however several obstacles to produce a portable and scalable Linda system ....

....shared by all processes joined in the computation. Linda was implemented in a wide range of platforms, with shared memory architectures, like the Encore, Sequent and Alliant machines as well as with distributed memory, like the Intel iPSC hypercube [38] Bell Labs S Net [8] and transputers[5][8]. There are also implementations in more loosely coupled systems, e.g. VAX VMS Workstation clusters [31] PC Networks [12] and Sun Workstations [24] In spite of its wide acceptance, there have been however several obstacles to produce a portable and scalable Linda system which can run efficiently ....

[Article contains additional citation context not shown here]

N.Carriero, D.Gelernter. "The S/Net's Linda Kernel", ACM Transactions on Computer Systems, Vol. 4, No. 2, pp.110-129, May 1986.

....joined in the computation. 1 PARIX is a registered trademark of Parsytec Computer Gmbh, Aachen, Germany. Linda was implemented in shared memory machines like the Encore, Sequent and Alliant machines as well as in distributed memory machines, like the Intel iPSC hypercube, Bell Labs S Net [8], and more loosely coupled systems, e.g. VAX VMS workstation clusters [31] PC networks [12] and Sun Workstations [24] It provides a collection of primitives for process creation and interprocess communication that can be added to existing sequential languages. Linda is nowadays a promising ....

....a Parsytec X plorer machine with eight nodes with the INMOS T805 and 4Mb RAM. The nodes are connected in a 4x2 grid. To assess the performance of programs written using ParLin we first decided to estimate the cost of a basic TS operation. For that purpose we used the well known pingpong program [8]. This program creates two processes ping and pong that stay in a loop making in( of ping tuples and out( of pong tuples, and vice versa. Table 1 summarizes the results obtained with ParLin along with results obtained with the same program for other Linda or Linda like implementations. Table 1: ....

N.Carriero, D.Gelernter. "The S/Net's Linda Kernel", ACM Transactions on Computer Systems, Vol. 4, No. 2, pp.110-129, May 1986.

....form a non null intersection with the replication space of the tuple requests. A trivial solution that follows this approach is one where the Out tuples are stored only on one node while the In and Read tuples search globally for the match and are eventually replicated on every node of the system [9]. At the other end of the spectrum, the dual solution provides a full replication of the Out tuples and a local search and a non replicated storage of the In and Read tuples. Both solutions, even if simple and systemindependent, are not effective for large system. In fact, they are based on a ....

N.Carriero, D.Gelernter, "The S/Nets's LINDA Kernel", ACM Transactions on Computer Systems, Vol. 7, No. 7, July 1985.

....The specific organization of kernels within one subPageSpace is a local decision. Known architectures for distributed implementation of Linda like systems include full replication of a repository to all nodes, no replication with a single, centralized repository, or a partial replication as in [20]. As long as there is one defined node that follows a gateway protocol to other sub PageSpaces, our architecture supports all of them. In fact, the current Jada implementation uses a centralized or fully replicated repository, whereas Laura implements a partial replication scheme. For a gateway, a ....

 N. Carriero and D. Gelernter, "The S/Net's Linda Kernel," ACM Trans. Computer Systems, vol. 4, no. 2, pp. 110--129, May 1986.

No context found.

Car86b. N. Carriero and D. Gelernter, "The S/Net's Linda Kernel," ACM Trans. Comp. Syst. 4(2), pp. 110-129 (May 1986).

No context found.

N. Carriero and D. Gelernter, "The S/Net's Linda Kernel," ACM Transactions on Computer Systems, pp. 110-129, May 1986.

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