Tanenbaum, Andrew and Woodhull, Albert 1997. Operating Systems: Design and Implementation, Prentice Hall.  @ NUS  Home/Search   Document Not in Database   Summary   Related Articles   Check

....for some jobs to be starved. In some sense FIFO is the optimally fair policy in that it optimizes the maximum flow time objective. Operating systems such as Unix don t implement pure SETF, or even pure MLF, the less preempting version of SETF. Once again this is out of fear of starving jobs [28, 29]. Unix systems adopt compromise policies that attempt to balance the competing demands of average QoS and fairness. In particular, Unix scheduling policies generally raise the priority of processes in the lower priority queues that are being starved [28] The desire to optimize for the average ....

....this is out of fear of starving jobs [28, 29] Unix systems adopt compromise policies that attempt to balance the competing demands of average QoS and fairness. In particular, Unix scheduling policies generally raise the priority of processes in the lower priority queues that are being starved [28]. The desire to optimize for the average and the desire to not have extreme outliers generally conflict. The most common way to compromise is to optimize the Lp norm, generally for something like p = 2 or p = 3. For example, the standard way to fit a line to collection of points is to pick the ....

[Article contains additional citation context not shown here]

A. Tanenbaum. Operating systems: design and implementation. Prentice-Hall, 2001.

....Before we describe the Tacoma paradigm, we will give a brief overview of the subject of process migration. This overview enables the reader to identify the main differences when compared to the moving agents paradigm. A process is one of the two major components of a distributed system. Tanenbaum [56] defines a process to be a program in execution, consisting of the executable program and all the other information, i.e. stack, pointers and registers, needed to run the program. The approach taken by the process migration community is to transfer this running process to other machines ....

Tanenbaum, Andrew S.: "Operating systems: Design and implementation", Prentice-Hall, 1987. 85

....effect of the cycle is the increase in cost of creating updating the RI. Cycle avoidance solution: In this solution we do not allow nodes to create an update connection to other nodes if such connection would create a cycle. The techniques for cycle avoidance have been extensively studied (see [22] for a survey) and we do not cover it further in this paper. The main disadvantage of this approach is that in the absence of global information we may end with a suboptimal solution. Cycle detection and recovery: This solution detects cycles sometime after they are formed and, after that, takes ....

A. Tanenbaum and A. Woodhull. Operating Systems Design and Implementation. Prentice-Hall, Inc., 1999.

....without any Operating System memory copy. To clarify the strong features of rDMA, consider the conventional distributed IPC based on message passing. These systems create OS support for complex bu#er management and a detailed protocol for send and receive acknowledgement and authentication [8, 117, 94]. The system overview is shown in figure B.6. The Operating System first copies the message into allocated bu#ers within the system space which is then transmitted across the network to the recipient s address space. The overhead for context switching and memory copying adds considerably to ....

Andrew S. Tanenbaum and Albert S. Woodhull. Operating Systems: Design and Implementation. Prentice Hall, second edition, 1997.

....implement a binary semaphore, one in which only a single process is allowed to hold the semaphore at a time. Spin locking implementations of binary semaphores have been extensively studied in the parallel computing and operating system communities. We point the reader to operating systems texts [68, 76], a survey by Dinning [23] and surveys with performance analysis by Anderson [5] and by Graunke and Thakkar [32] Spin locks are typically two orders of magnitude faster than operating system semaphores, and are generally acknowledged to be an important factor in the success of multi processor 6 ....

A. Tanenbaum, A. Woodhull, and A. Woodhull. Operating Systems: Design and Implementation. Prentice Hall, second edition, January 1997.

....above conventional operating systems. Implementors of persistent languages are invariably forced to construct an abstract machine above the operating system, since the components of a persistent system are different in nature to the components of a conventional operating system. For example, in [37], Tanenbaum lists the four major components of an operating system as being memory management, file system, input output and process management. In persistent systems, the file system and memory management components are unified. In many operating systems, input output is presented using the same ....

Tanenbaum, A. S. "Operating Systems: Design and Implementation", Prentice Hall, International Editions, 1987.

....operating systems. Existing operating systems do not provide an ideal platform for the development of persistent systems. This is not surprising since this was never part of their design goals. Indeed, most operating systems have files as their only abstraction over long term memory. Tanenbaum [38] has listed the four major components of an operating system as being memory management, file system, input output and process management. The nature of these four components is different in persistent systems. In a persistent system, the functionality of the file system and memory management are ....

Tanenbaum, A. S. "Operating Systems: Design and Implementation", International Editions, Prentice Hall, 0-13-637331-3, 1987.

....3.1. Previous Experiments using LSI In previous experiments done by the authors, the function and class declaration levels have been used [32] Two readily available sot tware systems were used as data for the experiments: LEDA [26] Library for Efficient Data structures and Algorithms) and MINIX [47] (Operating System) LEDA is a library of the data types and algorithms for combinatorial computing and provides a sizable collection of data types and algorithms in a form that allows them to be used by non experts. LEDA is composed of over 140 C classes. MINIX is a simple version of the UNIX ....

Tanenbaum, A. and Woodhull, A., Operating Systems Design andImplementation, Prentice Hall, 1997.

....above conventional operating systems. Implementors of persistent languages are invariably forced to construct an abstract machine above the operating system, since the components of a persistent system are different in nature to the components of a conventional operating system. For example, in [37], Tanenbaum lists the four major components of an operating system as being memory management, file system, input output and process management. In persistent systems, the file system and memory management components are unified. Some persistent systems require that the state of a process ....

Tanenbaum, A. S. "Operating Systems: Design and Implementation", International Editions, 0-13-637331-3, 1987. 14

....in the part dealing with output; the semantics part is modified only if the application needs new bi directional features. The bi directional formatters mentioned in the introduction have this property; other editors, e.g. emacs, have this property; and operating systems, such as UNIX[2] and MINIX[29] have this properry [21] Because this paper is about one bi directional editor, the topic of editors deserves a more thorough discussion here. The emacs[7, 30] editor is apparently also divided into an editing command processor part and a screen manager working from termcap information. It ....

A.S. Tanenbaum, Operating Systems: Design and Implementation, Prentice-Hall, Englewood- Cliffs, NJ, 1987.

....that it focuses on distributed computing on mobile computers, and it supports Windows NT 98 2000 ME as well as the CE PocketPC Handheld PC 2000 platforms. PortOS is a multi phase operating systems project designed to accompany a traditional presentation of operating systems found in OS textbooks [8, 10, 12]. These texts provide some insight into implementation issues, but mostly focus on theory. PortOS complements them with subprojects covering threads, scheduling, unreliable datagrams, reliable streams, ad hoc routing and file systems. Students program in C against a realistic, but sanitized, ....

....the project, as well typical figures for the amount of code the students write. 4 Related work PortOS has been used as the project for our honors level operating systems course. In this respect it is comparable to other instructional operating systems which have been described in the literature [3, 12]. We briefly elaborate the differences between PortOS and these systems, as well as other styles of operating system projects. q ne Figure 3 . A topology file and the corresponding broadcast network. quark .xx.x proton x.xx. neutron xx. electron .x. x neutrino x. x. neutron neutrino ....

[Article contains additional citation context not shown here]

Andrew S. Tanenbaum and Albert S. Woodhull. Operating Systems: Design and Implementation. Prentice Hall, second edition, 1997.

....depends (logarithmically) on the universe size, skip lists can also be favored over tries, especially when the universe size is large. There are many circumstances under which access patterns show considerable bias. It is well known that locality of reference results in high skew in memory cache [20], disk [14] and buffer [8] management for which a least recently used (LRU) policy yields good results. Recent studies on table lookups for Internet packet filtering and classification [11, 12, 5] demonstrate similar skew. In such applications, the access probability of a key k, decreases ....

Andrew Tanenbaum. Operating Systems: Design and Implementation, 2nd edition. Prentice-Hall, 1997.

.... software clock on the other hand, generally consists of a low grade quartz resonatorstabilised oscillator and a hardware counter that interrupts the processor at regular intervals known as the resolution, and adds a quantity known as a tick to a system variable representing the software clock time [9]. Software clocks will lose synchronism for a variety of reasons but principal among them is incorrect oscillator frequency. Oscillator frequency is also dependent on temperature which further complicates matters. Such clock oscillator errors can result in skew rates of a few hundred microseconds ....

Tanenbaum, A.,Woodhall, A.,Operating System Design and Implementation, 2nd Ed., Prentice Hall, 1997, pp. 222-227.

....questions arise as to whether the original memory has changed since the agent retrieved it, whether another agent is trying to read that information during the update, how do agents get notified that the information has changed, etc. Such issues are discussed in basic texts on operating systems [Tanenbaum, 1987] and computer architecture [Hwang and Briggs, 1985] My model avoids these problems by restricting the number of pieces of chalk, i.e. by providing a semaphore to the write process, the number of erasers, i.e. by providing a semaphore to the delete process, the amount of local memory an agent ....

A. Tanenbaum. Operating Systems: Design and Implementation. Prentice Hall, Engelwood Cliffs, NJ, 1987.

....and resumes program operation. The best performance is achieved if applications can run natively, i.e. outside an interpreter or virtual machine. Unfortunately, current processors do not support fine grained memory management, but rely on segmenting memory into pages of 4 or 8 KBytes [Tan87] This usually results in either coarsegrained access control, when multiple objects are put into the same page, or a waste of physical memory. Itzkovitz and Schuster [IS99] recently proposed a system allowing multiple objects to share the same physical page while keeping fine grained read write ....

Andrew S. Tanenbaum. Operating Systems: Design and Implementation. Prentice-Hall, 1987.

....and adds features such as hardware support for coherent shared memory, high end throughput of Gbyte s and good architectural support for design of multiprocessor systems. For computer communication, there exists two distinct methods of communication, message passing and shared memory [Tanenbaum87]. The idea of SCI LAN is closely connected to the distinction between these communication paradigms. All Local Area Networks of widespread use are based on message passing, but the advent of hardware supported coherent shared memory (implemented by transactions in the interconnect) allow computers ....

Tanenbaum, A., Operating Systems - Design and Implementation, Prentice Hall, 1987, ISBN 0-13-637331-3, pp. 51-70

....soft real time and millisecond response times [8] All of the currently available hard real time add ons depend on an additional realtime micro kernel, which runs a modified Linux kernel as a user process. Micro kernel architecture based operating systems do generally have lower response times [14, 15]. RTLinux is a good example of this approach (Figure 2(a) It is chronologically the first real time add on and it is currently enjoying the largest industrial support. The project goals are creating an operating system, capable of supporting both traditional Linux user processes and hard ....

Andrew Tanenbaum. Operating Systems: Design and Implementation. Prentice-Hall, 2nd edition, 1997.

....was decided to implement DIPC in the user space as much as possible, with minimal changes to the kernel. This can lead to a cleaner and simpler design, but in a monolithic operating system, such as Linux, has the drawback of requiring frequent copy operations between kernel and user address spaces [2]. As UNIX does not allow user space processes to access and change kernel data structures at will, DIPC has to have two parts: the more important part is a program named dipcd, which runs with superuser privileges. dipcd forks several processes to do its work. The other part is inside the kernel, ....

Andrew S. Tanenbaum, Operating Systems Design and Implementation, Prentice-Hall, 1987.

....simplified OS task scheduler and is represented with metaphors and 3 D visualization. 4.1. Step 1: Choose System to be Modeled The first step in the rube modeling methodology is Choose system to be modeled. In accordance with this step, a MINIX like OS is selected as the system to be modeled [41]. This OS can be structurally characterized by hierarchical, vertical layers (top to bottom) 41] user processes (e.g. word processor) server processes (e.g. file system) I O tasks (e.g. disk) and . process management (e.g. scheduling and communication) 10 J. F. HOPKINS AND P. ....

....1: Choose System to be Modeled The first step in the rube modeling methodology is Choose system to be modeled. In accordance with this step, a MINIX like OS is selected as the system to be modeled [41] This OS can be structurally characterized by hierarchical, vertical layers (top to bottom) [41]: user processes (e.g. word processor) server processes (e.g. file system) I O tasks (e.g. disk) and . process management (e.g. scheduling and communication) 10 J. F. HOPKINS AND P. A. FISHWICK A graphical representation of these layers can be seen in Fig. 2. The example model ....

A. S. Tanenbaum (1987) Operating Systems: Design and Implementation. Prentice-Hall, Englewood Cliffs, NJ, 940 pp.

....when an individual server cannot handle service requests because it is blocked making a request of its own. If a cycle in these client server dependencies exists, then the system is said to have deadlocked. To see why this is the case, we have to consider the four necessary conditions for deadlock [53], 17] from operating system theory: 1. Mutual exclusion condition Resources exist that are not sharable. 2. Non preemption condition Once a resource is given to a process, it cannot be revoked until the process voluntarily gives it up. 3. Hold and wait, or partial allocation condition ....

....are first assigned numbers. Although each resource is generally given a unique number, this is not a strict requirement for the algorithm to work. Using these assignments, if processes (or clients) always request resources (or servers) in increasing numerical order, then deadlock cannot occur [53]. The final approach for handling deadlock is called deadlock avoidance. In this technique, resources are not necessarily granted to a requesting process (or client) even if they are currently available, if by granting the resource places the system in a possible unsafe state. Typically, this ....

A. S. Tanenbaum, A. S. Woodhull, "Operating Systems: Design and Implementation ", Second Edition, Prentice-Hall Inc., Upper Saddle River, New Jersey, 1997.

....an application. 6 can be with any input output channel which include storage device access. I O channels connect external servers and clients executing in parallel, local processes executing concurrently, local data files, and local virtual devices provided by the native operating system. See [TANE 87] for the respective descriptions of sockets, pipes, file descriptors, and devices. Communication via storage device access is achieved by one user writing to the memory thereby allowing another user to access that memory. 2.4 Types of Entities Every device of the underlying operating system ....

Tanenbaum, A.S., Operating Systems: Design and Implementation, Prentice Hall, Englewood Cliffs, NJ, 1987.

....and are likely to be accessed again in the future By using a cache, when a page is to be read written, then we first check whether it is in the cache and if it is, the operation can be completed in memory and we say that there was a cache hit. For a general description of caching techniques, see [7]. In the rest of this section, we propose specific methods for clustering and caching when storing arbitrary graphs. Section 2.3 describes a naive file structure for graph storage, which is improved upon in Section 2.4. Section 2.5 describes the incremental clustering mechanism and Section 2.6 ....

Andrew S. Tanenbaum. Operating Systems: Design and Implementation. PrenticeHall, Inc., 1987.

....algorithm, the system memory usage can be bounded by a least recently used (LRU) cache management strategy that reuses memory for linetrees and interpolants. The interpolant ray tracer implements a linetree cache management algorithm similar to the UNIX clock algorithm for page replacement [Tan87] though it manages memory at the granularity of linetree cells rather than at page granularity. The system allocates memory for linetrees and interpolants in large blocks. When the system memory usage exceeds some user specified maximum block count, the cache management algorithm scans through ....

A. S. Tanenbaum. Operating Systems: Design and Implementation. Prentice-Hall, Inc., 1987. 167

....networks can contribute to increased understanding of message passing systems and their dynamic behaviour. Before presenting the examples, we give brief introductions to MINIX and to interaction networks. MINIX MINIX is a complete operating system developed for use in operating systems teaching [4, 5]. While the user and application programmer interfaces are heavily based on Unix, the internal structure of MINIX is quite different to the monolithic structure of traditional Unix kernels. In MINIX, the operating system consists of three programs and around ten processes. The kernel program ....

....that are made up of communicating processes executing more than one program. We are not aware of other monitors that can do this. Tanenbaum seems to regard graphical displays of message passing patterns as a useful teaching device, as he uses them to explain MINIX message passing patterns [4]. In his figures, each process is represented by a single vertex, with each message shown as an edge. The edges are numbered to show the order in which the messages are sent. The figures presented in [4] rely on the availability of a MINIX expert to produce them, whereas an interaction network ....

[Article contains additional citation context not shown here]

TANENBAUM,A.S.Operating Systems: Design and Implementation. Prentice-Hall, Englewood Cliffs, NJ, 1987.

....is robust with respect to its performance improvement. The above analysis was based on the worst case assumption that a full rotation is made during every subscriber switch. On the average, however, we can reasonably expect that the rotation time to access the first block is only (l rot max =2) [Tane87]. Using this value instead of l rot max in the numerator of Equation (6) means that, in the typical case, the memory usage factor of SCAN is in the interval (0; 0:25) relative to the QUAD scheduler. In a general computer system environment, where user requests may arrive in random fashion, ....

Tanenbaum, A.S., "Operating Systems: Design and Implementation", Prentice-Hall, 1987.

....operating systems. Existing operating systems do not provide an ideal platform for the development of persistent systems. This is not surprising since this was never part of their design goals. Indeed, most operating systems have files as their only abstraction over long term memory. Tanenbaum [44] has listed the four major components of an operating system as being memory management, file system, input output and process management. The nature of these four components is different in persistent systems. In a persistent system, the functionality of the file system and memory management are ....

A.S. Tanenbaum, "Operating Systems: Design and Implementation", International Editions, Prentice Hall, 1987.

....the communication 10 structure. This approach is well suited to high message rates or bursty traffic, such that the sender does not have to hold a message too long before passing it further. It has been successfully applied in operating systems which consist of several interacting components [TW97, CRJ87] A consequence of the Payloads pattern is that adding new message types does not require changing the existing entities which are not interested in them. For example, in a Data flow architecture (Section 2) filters pass downstream the messages they do not understand, without performing ....

Andrew S. Tanenbaum and Albert S. Woodhul. Operating Systems---Design and Implementation. Prentice-Hall, second edition, 1997. ISBN 0-13-638677-6.

....is unlikely that the problem can be effectively solved completely at any one level. 1.2 Virtual Memory Management and the Page Placement Policy The strategy employed by an operating system to manage virtual memory is characterized by the paging policy. Operating systems textbooks (e.g. CD73, Tan87, PS85] typically discuss paging policies in terms of three subpolicies. The fetch policy determines which page will be brought into main memory and when that page fetch will be initiated, the placement policy determines which empty frame will be used to hold the page being fetched, and the ....

A. S. Tanenbaum. Operating Systems - Design and Implementation. Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1987.

.... event . 133 event: statechange ; predicate ; predicate: tempexpr eventlistopt . eventlistopt: eventlist . 134 Appendix B Interactive Session with the Slicer The following is a typescript of a session with the Tester s Assistant. The program used in this run is the Minix [62] login program. The program has already been run through the C pre processor. Several interesting library functions have been specified from the command line as the Tester s Assistant program ta.exe is invoked. Several indicators as to the size of the graph are printed as analysis is complete. At ....

Andrew S. Tanenbaum. Operating Systems -- Design and Implementation. Prentice--Hall, 1987.

No context found.

Tanenbaum, Andrew and Woodhull, Albert 1997. Operating Systems: Design and Implementation, Prentice Hall.

No context found.

A. S. Tanenbaum and A. S. Woodhull, Operating Systems: Design and Implementation. Prentice-Hall, 1997.

No context found.

Andrew S. Tanenbaum, Albert S. Woodhull. Operating Systems Design and Implementation, 2nd Edition. Prentice Hall, 1997.

No context found.

Tanenbaum, A. S. (1987). Operating Systems: Design and Implementation. Prentice Hall, Englewood Cliffs,1989.

No context found.

Tanenbaum, A.S., Operating Systems: Design and Implementation, Prentice Hall, 1986.

No context found.

A. S. Tanenbaum and A. S. Woodhull. Operating Systems -- Design and Implementation. Prentice-Hall, second edition, 1997.

No context found.

A. Tanenbaum. Operating systems: design and implementation. Prentice-Hall, 2001.

No context found.

A. Tanenbaum. Operating systems: design and implementation. Prentice-Hall, 2001.

No context found.

A. Tanenbaum. Operating Systems: Design and Implementation. Prentice Hall, 2nd edition, 1996. 5

No context found.

TANENBAUM, A. S. & WOODHULL, A. S. Operating systems: design and implementation. 2nd. Ed., Prentice-Hall , 1997.

No context found.

A. S. Tanenbaum. Operating Systems: Design and Implementation. Prentice-Hall, Inc., 1987.

No context found.

A. Tanenbaum, A. Woodhull, "Operating Systems: Design and Implementation", Prentice Hall Inc., 1997, ISBN 0-13-638677-6

No context found.

A. S. Tanenbaum and A. S. Woodhull, Operating Systems: Design and Implementation. Prentice-Hall, 1997.

No context found.

Tanenbaum & Woodhull. Operating Systems: Design and Implementation. Prentice Hall 1997

No context found.

A.S. Tanenbaum, Operating Systems: Design and Implementation, Prentice-Hall, 1987.

No context found.

Andrew S. Tanenbaum, Albert S. Woodhull, "Operating Systems: Design and Implementation", Prentice Hall, 2 nd edition, 1997

No context found.

Tanenbaum, A.S., Woodhull, A.S., "Operating Systems, Design and Implementation". Prentice-Hall, ISBN 0-13-638677-6, 2nd ed., 1997.

No context found.

A. S. Tanenbaum. Operating systems: design and implementation. Prentice-Hall, Englewood Cli#s, NJ, 1987.

No context found.

A. Tanenbaum. Operating Systems Design and Implementation. Prentice Hall, Inc., 1987.

No context found.

Andrew S. Tanenbaum, Albert S. Woodhull. Operating Systems Design and Implementation. Prentice-Hall, Inc. 1997

No context found.

Andrew S. Tanenbaum "Operating Systems: Design and Implementation"

First 50 documents  Next 50

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

CiteSeer.IST at NUS - Copyright Penn State and NEC. Hosted by the School of Computing, National University of Singapore.