Cheriton, C.R., "The V Distributed System", CACM, V31.3, 1988.  Home/Search   Document Not in Database   Summary   ACM   TOC   Related Articles   Check

....forms of migration mechanisms could be employed in different situations to handle load sharing. Current distributed computing systems facilitate some forms of migration, e.g. data, process, or object migration, aiming to achieve better system performance. For example, the Galaxy [15] and the V [4] Distributed Systems support process migration. Mach supports task migration [2] and Emerald supports object migration [9] The process and object migrations are attractive mechanisms for load sharing. However, since the heterogeneity of a distributed system is increased with the joining and ....

V.R. Cheriton, "The V Distributed System," Communications of the ACM, vol. 13, no. 3, pp.314-333, Mar. 1988.

....kernel threads running in the same user process share the same address space, they are scheduled independently by the kernel scheduler. Each scheduling requires the transfer of control between the kernel level and the user level, still causing too much overhead. Mach[3, 4] Topaz[1, 9] and V[10] support the kernel thread mechanism. The other one is the user level thread. User level threads are managed by runtime routines pre compiled and linked to the user application as a library. User level execution is transparent to the kernel. Scheduling among the userlevel threads does not require ....

D. Cheriton, "The V Distributed System," in Commun. ACM, vol 31, no. 3, pp. 314-333, Mar 1988.

....networks and implementing reliability issues such as atomic transactions and replication. 48 Contact: David R. Cheriton or Tony Mason, Computer Science Department, Stanford University, Stanford CA 94305. References: 318] 319] 320] 321] 322] 323] 324] 325] 326] 327] 328] [329], 330] 331] 332] 333] 146] 334] 29] 335] 336] 2.51 VAXcluster Main Goal The Digital Equipment Corporation has been developed the VAXcluster product as a set of closely coupled DEC VAX computers providing a highly available and extensible configuration that operates access , ....

D. Cheriton, "The V Distributed System", Communications of the ACM, 31(3):314--333, March 1988.

....literature on reducing data touching overheads. Such efforts include protocol integrated layer processing [2, 5, 11] high performance network adapters to eliminate copying between devices and the operating system kernel [6] and restructuring operating system software to minimize data movement [4, 6, 8, 9, 15, 17, 18]. Much of the previous work in this area concentrated on reducing data copying for network protocol processing, particularly for the TCP network protocol [16] We are interested in providing an I O framework for the UNIX operating system that has the following characteristics: The framework ....

....but to our knowledge, there is no one system that meets all of our requirements. For example, there is a lot of work in the area of reducing network protocol processing that, in general, is not applicable to other kinds of I O [2, 5] Some systems that implement a zero copy mechanism such as [4, 9] do not meet our requirement for efficient support on MP systems. The promising work of Druschel and Peterson is mostly geared to network protocols in a non UNIX environment [8] Finally, most of the previous systems do not address compatibility with existing Streams modules. In this paper, we ....

[Article contains additional citation context not shown here]

Cheriton, D.R. "The V Distributed System." Communications of the ACM, vol. 31, no. 3 (March 1988).

....investment to protect. Further, for broader market appeal we felt it unwise to rely on specialized hardware, as some of the cited examples in literature did. We turned to virtual memory remapping and copy onwrite technique, both being widely adopted in operating system design to avoid copying [1, 2, 8, 16]. Although these operations are not without expense, they often can be applied transparently, and so fit our goal of minimizing software module change. Networking adaptors used are Sun s SBus based ATM interface cards, capable of both 155 and 622 Mb s, OC 3 and OC 12 respectively. In the next ....

D. R. Cheriton. "The V distributed system," Communications of the ACM, vol.31, no.3, March 1988.

....different site during execution. In this way, processes can be moved from any site so that the load can be maintained in a balanced state. Distributing the load using process migration is called load balancing. Examples of operating systems that provide process migration include Charlotte [12] V [13], Accent [14] and Sprite [1] Several load sharing algorithms have been developed. Most of these algorithms can be implemented as load balancing algorithms, if the targeted operating system is one that supports process migration. Load sharing is simpler, but load balancing provides more ....

D. R. Cheriton, "The V distributed system," Communications of the ACM, vol. 31, pp. 314--333, March 1988.

....as threads, address spaces, and inter process communication. These abstractions can be composed to build user level servers that implement common services such as file systems, processes, and virtual memory. Examples of micro kernel architectures include Mach [1] Chorus [12] Amoeba [35, 45] V [10], Choices [9] Psyche [42] L3 L4 [28, 29] and Arena [34] The most notable effect of the micro kernel design is to separate the implementation of the system into smaller, more manageable pieces in the form of the kernel and the user level servers. This arrangement greatly improves the ....

D.R. Cheriton, "The V Distributed System", Communications of the ACM, 31(3), pp. 314-333, 1988.

....(not kernel to kernel) and do not involve any tricks or special cases. Noteworthy are the performance of the V system [15,16] and the Firefly RPC [17] V because it is widely thought to be the fastest distributed system currently in existence and the Firefly because it is a multiprocessor. For V [18] we find that delay for null RPCs is 2.54 msec (vs. 1.4 msec for Amoeba) and that the bandwidth for 8 Kbyte RPCs is 460 Kbytes sec (vs. 625 Kbytes sec for Amoeba) For larger requests (up to 16 Kbytes in [18] the data rate increases to 550 Kbytes sec (vs. 644 Kbytes sec for Amoeba) All of ....

....system currently in existence and the Firefly because it is a multiprocessor. For V [18] we find that delay for null RPCs is 2.54 msec (vs. 1.4 msec for Amoeba) and that the bandwidth for 8 Kbyte RPCs is 460 Kbytes sec (vs. 625 Kbytes sec for Amoeba) For larger requests (up to 16 Kbytes in [18]) the data rate increases to 550 Kbytes sec (vs. 644 Kbytes sec for Amoeba) All of these figures represent a performance considerably worse than that of Amoeba, despite the fact that the V measurements were made on substantially faster hardware, namely SUN 3 75s (vs. SUN 3 50s for Amoeba) The ....

Cheriton, D.R.: "The V Distributed System," Commun. ACM , vol. 31, no. 3, pp. 314333, Mar. 1988.

....operating system compatible with UNIX. Its basic design goal was the development of new technologies for the implementation of UNIX like facilities in a workstation cluster [11] 12] assuming a traditional model of computation. V Distributed Operating System The V distributed system [16] is an operating system for a cluster of workstations connected via a high speed intercommunication network. MOSIX MOSIX [17] is a distributed operating system that allows to enhance the utilization of a collection of workstations and that makes them appear as a single machine UNIX ....

Cheriton D.R.: "The V Distributed System", Communication of the ACM, 31(3)314-333, March 1988.

....the interface to the host Sun workstation. The remaining links can be reconfigured to provide fast local communication in a topology tuned to a particular application. 1.2. ARTiSt ARTiSt (A Run TIme SysTem) is a software backplane similar in approach to the distributed operating systems V [2], Amoeba [3] and Helios [4] and provides distributed runtime support for multiple tasks on a transputer network. Each node runs an instance of the ARTiSt kernel which provides topology transparent communication primitives. Tasks communicate via variable length messages sent through the ARTiSt ....

D.R. Cheriton, "The V Distributed System" CACM 31: 314-333, 1988.

....on the underlying network providing a broadcast mechanism at the data link layer for good performance. 6. 4 V System The V system distributed operating system was developed at Stanford University, California, as a platform for the development of distributed operating system designs and concepts [Cheriton 1988]. In V system, the group communication protocol has been integrated with the RPC primitive to provide basic interprocess communication for the operating system [Cheriton and Zwaenepoel 1985] In V system, a RPC call is issued by the client process allowing the client process to send a service ....

Cheriton D.R. "The V Distributed System", Communications of the ACM, March, 314-331.

.... simplified (because request data can be left in a client s buffer, and the response data can be stored directly in this buffer) and the transport level protocol is simplified (because error handling as well as flow control exploit the response to acknowledge a request and authorise a new request) [Cheriton 1988]. When the requesting process is blocked waiting for a reply, it can be blocked indefinitely. This can occur because of a communication failure, failure of a destination computer, or simply because the server process does not exist any longer or is too busy to compute a response in a reasonable ....

....RPC because there are no standards for messages. Moreover, neglecting language aspects of RPC and because of the variety of message passing semantics, these two facilities can look very similar. Examples of a message passing system that look like RPC are message passing for the V system (which in [Cheriton 1988] is called now the remote procedure call system) and message passing for Amoeba [Tanenbaum and van Renessee 1985] and RHODOS [De Paoli et al. 1995] Comparing the remote procedure call and message passing, the former has the important advantage that the interface of a remote service can be easily ....

[Article contains additional citation context not shown here]

D. R. Cheriton, "The V Distributed System", Communications of the ACM, March, 314-331.

....resources. The introduction of local area networks in the 1980 s expanded this role even further. A goal of network operating systems such as Locus [Popek CHAPTER 8. RELATED WORK 168 et al. 1981] Mach [Accetta et al. 1986] Sprite [Nelson et al. 1988] Amoeba [Mullender et al. 1990] and V [Cheriton 1988] was to make remote resources over the LAN as easy to use as local resources, in the process simplifying the development of distributed applications. For example, distributed operating systems provide abstractions such as communication channels and processes, as opposed to networks and processors. ....

D. R. Cheriton. "The V Distributed System". In Communications of the ACM, pp. 314--333, March 1988.

....executing on massively parallel environments. This is because each processing unit must communicate with another over a network which has a higher cost of communication [CAP93] Some distributed environments s are Berkeley s NOW [CULL97] MOSIX [BARA98] Amoeba [TANE91] Sprite [OUST88] V [CHER88], Chorus [ROZI92] Globe [STEE97] Beowulf [RIDG97] and RHODOS [DePA95] The type of parallel execution that needs to be performed will most certainly play an important part in determining what type of parallel execution environment is chosen. The particular application may be either computation ....

D.R. Cheriton. "The V Distributed System", Communications of the ACM, Vol. 31, No. 3, pp314-333, March 1988.

....spaces. Existing thread mechanisms, however, make unnecessary vertical and horizontal switchings on multiprogrammed systems, primarily due to inappropriate interfaces between the kernel and the user level. In most conventional thread mechanisms, threads are implemented solely by the kernel [15, 4, 16, 12] or by user level programs on top of the kernel level threads support [5, 1, 14, 6, 2, 7, 8, 11, 13] let us call threads implemented by the kernel k threads, and those implemented by user level programs on top of k threads u threads) These thread mechanisms share two characteristics that cause ....

D. R. Cheriton, "The V Distributed System," Communications of the ACM, vol. 31, pp. 314--333, Mar. 1988.

....useful for load balancing, software fault tolerance and adaption to changes in available communication resources etc. 93 94 It can be easily seen that changes in geometry involve only implementation issues. Geometric reconfiguration has been considered in the form of process migration in [AF89, Che88] These implementations support the migration of a process to another processor of the same architecture and running the operating system. The Polylith software bus system [PH91] supports geometric reconfiguration across processors of different architectures. Architecture independence is achieved ....

D. Cheriton. "The V distributed system". Communications of the ACM, pages 314--333, 1988.

....Xerox corporate environment, is another example of a transnational distributed system. Other examples of distributed operating system projects include Sprite [42] at the University of California, Berkeley, Eden [43, 44] at the University of Washington, Mach [45] at Carnegie Mellon University, V [46] at Stanford University, Locus [47] at the University of California, Los Angeles, and Clouds [48] at the Georgia Institute of Technology. Other projects in distributed systems have evolved algorithms for distributed object management [49, 50, 51] 1.3 The X Window System The X Window System [12, ....

D. R. Cheriton, "The V distributed system," Communications of the ACM, vol. 31, pp. 314--333, March 1988.

....like to thank my brother, Alexander Kamb, for his generous support and encouragement during my graduate studies. CHAPTER 1 INTRODUCTION Remote communication is a fundamental element of distributed systems. In operating systems created from scratch to support distribution, such as Amoeba and V [2, 3], the remote communication mechanism is a cornerstone of their development. Other systems have been extended to support remote communication, either explicitly as with sockets in Unix, or by transparent extension of the local interprocess communication (IPC) mechanism, as in Mach [4] With the ....

D. R. Cheriton, "The V distributed system," Communications of the ACM, pp. 314--333, March 1988.

No context found.

Cheriton, C.R., "The V Distributed System", CACM, V31.3, 1988.

No context found.

D. Cheriton, "The V Distributed System", in Communications of the ACM, vol 31, no. 3, pp. 314-333, Mar 1988.

No context found.

D. R. Cheriton, `The V distributed system', Communications of the ACM, 31, (3), 314--333 (1988).

No context found.

D. R. Cheriton, `The V distributed system', Communications of the ACM, 31, (3), 314--333 (1988).

No context found.

D. R. Cheriton, `The V distributed system', Comm. ACM, 31, (3), 314--333 (1988).

No context found.

D.R. Cheriton, "The V Distributed System," CACM. 31,3, March 1988, pp.314-333

No context found.

D. Cheriton. "The V Distributed System". Communications of the ACM, 31(3), 314-33.

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