R. Clark, D.E. Jensen, and F.D Reynolds, An Architectural Overview of the Alpha Realtime Distributed Kernel, In Proceedings of the USENIX Workshop, In 1993.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....threads. We do not deviate from the RMI programming paradigm that implies different semantics for argument passing in local and remote method invocations. 1. 1 Related work Distributed threads are a well known concept introduced in the context of the Alpha distributed real time operating system [4]. A distributed thread is a light weight process that spans multiple address spaces. In Alpha, a distributed thread is associated additionally with a set of parameters and attributes for scheduling and real time behavior, independent of the node where the current head of the control flow is just ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An architectural overview of the Alpha Real-Time Distributed Kernel. In Proceedings of the USENIX Workshop on Microkernels and other Kernel Architectures, pages 127--146. USENIX, 1992.

....create such distributed threads, but it is neither aware of them nor does it provide any means to control them. This problem is tackled in our contribution. 1. 1 Related work Distributed threads are a well known concept introduced in the context of the Alpha distributed real time operating system [3]. A distributed thread is a light weight process that spans multiple address spaces. In Alpha, a distributed thread is associated with a set of parameters and attributes for scheduling and real time behavior, independent of the 1 node where the current head of the control flow is just executing. ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An architectural overview of the Alpha Real-Time Distributed Kernel. In USENIX Workshop on Microkernels and other Kernel Architectures, pages 200--208, 1993.

....Fig. 2. Context per JVM Thread 2.2 Distributed thread identity to the rescue A Java program is executed by means of a JVM thread. Such a thread is the unit of computation. It is a sequential flow of control within a single address space, i.e. JVM. We introduce the notion of distributed thread [3] as a logical sequential flow of control that may span several address spaces, i.e. JVMs. A distributed thread is physically implemented as a concatenation of local JVM threads, sequentially performing remote method invocations when they transit JVM boundaries. For example, in Figure 2 the ....

....transformation process then becomes a relevant performance factor. 5 Related work We discuss related work according to the related fields that touches our work. Distributed threads. D. Jensen at CMU already introduced the notion of distributed thread in the Alpha distributed real time OS kernel [3]. The main goal of distributed threads in the Alpha kernel was integrated end to end resource management based on propagation of scheduling parameters such as priority and time constraints. In our project we adopted the notion of distributed thread at the application level. This allows the ....

R Clark, D.E. Jensen, and F.D Reynolds, "An Architectural Overview of the Alpha Real-time Distributed Kernel", Proceedings of the USENIX Workshop, 1992.

....overhead induced by the transformation process which is not small for the current implementation becomes a relevant performance factor. 7 Related work The discussion on related work is organized according to the related fields that our work touches. 7. 1 Distributed Threads Existing work [4] in the domain of (real time) distributed operating systems has already identified the notion of distributed thread as a powerful basis for solving distributed resource management problems. D. Jensen at CMU introduced the notion of distributed thread in the Alpha distributed real time OS kernel. ....

R Clark, D.E. Jensen, and F.D Reynolds, "An Architectural Overview of the Alpha Realtime Distributed Kernel", In Proceedings of the USENIX Workshop, April 1992.

....induced by the transformation process which is not small for the current implementation becomes a relevant performance factor. 21 7. Related work The discussion on related work is organized according to the related fields that our work touches. 7. 1 Distributed Threads Existing work [4] in the domain of (real time) distributed operating systems has already identified the notion of distributed thread as a powerful basis for solving distributed resource management problems. D. Jensen at CMU introduced the notion of distributed thread in the Alpha distributed real time OS kernel. ....

R Clark, D.E. Jensen, and F.D Reynolds, "An Architectural Overview of the Alpha Real-time Distributed Kernel", In Proceedings of the USENIX Workshop, April 1992.

....in the jukebox. This new kernel driver is almost identical to a standard SCSI disk driver [9] 1.3 Advantages of a Modular Approach The modular approach has three major advantages. The first is that less time is spent in the kernel, which has been shown to improve the kernel response time [6, 4]. The second advantage is that it is possible to customise the interface between an executing application and the underlying disks to improve performance. Only those modules required need be attached. For example, in experiments on databases, where exact placement of data on disk is important the ....

Clark, R., Jensen, E., Reynolds, F. "An Architectural Overview Of The Alpha Real-Time Distributed Kernel", Usenix Workshop on Microkernels and Other Kernel Architectures, April 1992.

....systems, services are more con cerned with maximizing overall utility (by serving the most important request first) than guaranteeing reserved resources for individual requests. Priority driven services can generally be categorized this way, and are supported in real time kernels such as Al pha [8] and Mach [9] Under overload conditions, lower priority tasks are denied service in favor of more im portant tasks. In the Rialto operating system [10] a resource planner attempts to dynamically maximize user perceived utility of the entire system. However, the scheme does not adopt the notion ....

R. Clark, E. Jensen, and F. Reynolds, "An ar- chitectural overview of the Alpha real-time distributed kernel," in Proceedings of the USENIX Workshop on Microkernels and other Kernel Architectures, 1992.

....execute as flows of control that may cross physical node boundaries, contrary to how conventional Java threads are confined to a single address space. In the remainder of this paper we refer to such a distributed computational entity as a distributed thread of control, in short distributed thread [2]. A distributed thread is a logical sequential flow of control that may span several address spaces (i.e. JVMs) As shown in Figure 1, a distributed thread # is physically implemented as a concatenation of local (per JVM) threads [t1, t4] sequentially performing remote method invocations when ....

....we integrated distributed thread identity into the byte code transformation scheme and solved this problem in an elegant way. 3. DISTRIBUTED THREADS AND DISTRIBUTED THREAD IDENTITY The notion of distributed thread has already been introduced in the Alpha distributed real time OS kernel at CMU [2]. D. Jensen has identified the notion of distributed thread as a powerful basis for solving distributed resource management problems. We borrow the definition of distributed threads from this work: A distributed thread is the locus of control point movement among objects via operation ....

R Clark, D.E. Jensen, and F.D Reynolds, "An Architectural Overview of the Alpha Real-time Distributed Kernel", in Proceedings of the USENIX Workshop on Microkernel and Other Kernel Architectures, April 1992.

....execute as flows of control that may cross physical node boundaries, contrary to how conventional Java threads are confined to a single address space. In the remainder of this paper we refer to such a distributed computational entity as a distributed thread of control, in short distributed thread [2]. A distributed thread is a logical sequential flow of control that may span several address spaces (i.e. Java Virtual Machines (JVMs) As shown in Figure 1, a distributed thread t is physically implemented as a concatenation of local (per JVM) threads 1 A local program executes completely ....

....the SCS. However from a control flow point of view, the threads t1 and t3 logically belong to the same distributed thread. The JVM monitor manager is however not aware of this logical connection between threads t1 and t3. 3. DISTRIBUTED THREADS AND DISTRIBUTED THREAD IDENTITY Existing work[2] in the domain of (real time) distributed operating systems has already identified the notion of distributed thread has as a powerful basis for solving distributed resource management problems. We borrow the definition of distributed threads from this work: A distributed thread is the locus of ....

[Article contains additional citation context not shown here]

R Clark, D.E. Jensen, and F.D Reynolds, "An Architectural Overview of the Alpha Real-time Distributed Kernel", in Proceedings of the USENIX Workshop on Microkernel and Other Kernel Architectures, April 1992.

....system is underloaded and chooses the subset of tasks that maximize the value of the computation per unit of time (value density) 3 when the system is overloaded. The Alpha effort [12] introduced the concept of time valued functions, which associate a value according to the task finishing time [7]. The function presents a drop in value after the deadline has passed, and beyond a certain time the value drops to zero. Alpha provides data replication as well as migration for fault tolerance, without explicitly addressing their effect in the timing constraints of applications or the effect of ....

R.K.Clark, E.D. Jensen, F.D. Reynolds. "An Architectural Overview of the Alpha Real-Time Distributed Kernel", USENIX Workshop on Microkernels and other Kernel Architectures, 1992.

....Work on Operating Systems More recently, resource reservation has been applied for temporal isolation of real time applications [63, 75, 88] in an attempt to safely colocate real time and best effort processing via appropriate operating system support. While earlier real time operating systems [39, 138] provided a priority based interface, new kernel extensions have been proposed to provide real time guarantees for QoS sensitive applications. For example, capacity reserves [88] have been used in Mach to allocate processing capacity for multimedia applications [75] and flexible CPU reservation ....

....the state of the art in the provision154 ing of end to end QoS. Reservation based protocols were suggested to provide QoS guarantees for communication services [18, 65] Support for QoS or preferential service in the network has been examined integrated and differentiated services on the Internet [24, 26, 39, 144]. Several classes of service were considered, including guaranteed service which provides guaranteed delay bounds, and controlled load service which has more relaxed QoS requirements. Issues involved in sharing link bandwidth across multiple classes of traffic are explored in [46] The signaling ....

R. Clark, E. Jensen, and F. Reynolds, "An architectural overview of the Alpha real-time distributed kernel," in Proceedings of the USENIX Workshop on Microkernels and other Kernel Architectures, 1992.

.... pull for the decentralization of asynchronous real time systems that is both involuntary and voluntary. The most common involuntary motivation for decentralization is that the assets of the application (e.g. the radars and missile launching devices of a combat system) are inherently dispersed [2]. Furthermore, real time (response time) requirements of individual components of such systems often, cannot be met with a centralized computing facility. A primary voluntary reason for decentralization is survivability, in the sense of continued availability with a degradation of functionality or ....

....real time (response time) requirements of individual components of such systems often, cannot be met with a centralized computing facility. A primary voluntary reason for decentralization is survivability, in the sense of continued availability with a degradation of functionality or performance [2]. Often, it may be cost effective to physically distribute a mission management system than it is to implement as a centralized system that becomes a single point of failure. 4. A Benchmark Adaptive Resource Management Problem To illustrate how feedback control laws can be constructed for ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds, "An Architectural Overview of the Alpha Real-Time Distributed Kernel," Proceedings of The USENIX Workshop on Microkernel and Other Kernel Architectures, Seattle, April 1992.

....QoS. Additionally, processor capacity reserves do not aim at providing di erentiated QoS in clusters, nor do they provide predictable QoS in application level metrics in server systems. 5.1. 7 Migrating threads in Mach and Shuttles in Spring The migrating threads [46] of Mach [1, 23] and AlphaOS [32], and the shuttles of Spring [56] allow the resource consumption of a thread (or a shuttle) to be accounted towards the correct resource principal, when the thread (or shuttle) moves across 99 protection domains. However, a single thread cannot perform work on behalf of several resource ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An architectural overview of the alpha real-time distributed kernel. In Workshop on Micro-Kernels and Other Kernel Architectures, pages 127-146, Seattle, WA, Apr. 1992. USENIX.

....any number (including zero) of loci executing within it. Parameters consisting of data and capabilities for other containers (see below) may be passed on an invoke operation. This mechanism is very similar to those provided in other object based systems such as Monads [14] Clouds [6] Alpha [5] and Spring [9] One of the key features to the invocation mechanism is that it is potentially very fast in comparison to message passing systems[2, 10] and for this reason a variant of this model has been incorporated into Mach[8] 2.3 Mapping and address space composition The purpose of ....

Clark, R. K., Jensen, E. D. and Reynolds, F. D. "An Architectural Overview of the Alpha Real-Time Distributed Kernel", Usenix Workshop on Microkernels and Other Kernel Architectures, 1992.

....over another. Many schedulers are including more attributes to better differentiate the relative importance of one task compared to others. The real time operating system Alpha is based upon the Benefit Accrual Model, which captures the system benefits of honoring particular timing constraints [CJR92] This allows the scheduler to make more intelligent decisions where tradeoffs need to be considered. Many real time systems use rate monotonic (RM) scheduling because it is very simple to implement. Static priorities are assigned to tasks based on their periods, where a shorter period implies a ....

....explicitly check to see if the timeout was exceeded. Another approach is to allow only asynchronous communication which simplifies execution time estimates. However, reasoning about the correctness of the program in the presence of concurrent activity and exception handling is much more difficult [CJR92] In multiple processor systems, a real time network is required to have bounded latencies so that schedulers can make appropriate decisions. These latencies can be affected by the choice of communication protocol. Selecting a broadcast or point to point protocol may depend on the topology of ....

Raymond K. Clark, E. Douglass Jensen, and Franklin D. Reynolds. An architectural overview of the Alpha real-time distributed kernel. In Proceedings of the USENIX Workshop on Micro-Kernels and Other Kernel Architectures, pages 127--146, April 1992.

....which resource consumptions are actually charged to the owner. In Stride and Lottery scheduling [Waldspurger 95] resource rights are encapsulated by tickets. Tickets are owned by clients, and ticket transfers allow to transfer resource rights between clients. In Mach [Ford et al. 94] AlphaOS [Clark et al. 92] and Spring [Hamilton et al. 93] migrating threads respectively shuttles correspond to threads and own resources. Migration of threads between protection domains enables these systems to account resource consumption of independent activities to the correct owner, i.e. thread. In [Ford et al. ....

Clark, R.K., Jensen, E.D., Reynolds, F.D.: An Architectural Overview of the Alpha Real-Time Distributed Kernel, Workshop on Micro-Kernels and other Kernel Architectures, April 1992

....a real time scheduler; and a method for dynamically enforcing real time guarantees. Additional details are discussed in [9] There are a number of systems that have components related to those described in this paper. The predominant ones are RT Mach [11] the Spring Kernel [6] the Alpha Kernel [2], the Rialto operating system [3] Mercer s processor capacity model [5] and the continuous media model [1] Real time processes are processes where the correctness of the process is determined not only by the actions of the process, but also the time at which these actions occur. The Dreams ....

Clark, R.K., Jensen, E.D., and Reynolds, F.D. An Architectural Overview of the Alpha Real-Time Distributed Kernel. In Proceedings of the USENIX Workshop on Microkernels and Other Kernel Architectures (Apr. 27-28, Seattle). 1992.

....by resource containers is similar to that provided by a number of abstractions developed in the context of multimedia and real time operating systems. These include the processor capacity reserves of Mercer et al. 22] the activities [18] of Rialto, the migrating threads of Mach [14] and AlphaOS [11], and the shuttles of Spring [16] The chief differences are related to the more general nature of resource containers, and the fact that none of the above abstractions addresses the problem of controlling resource usage of kernel I O processing. Almeida et al. 1] attempted to implement QoS ....

R. K. Clark et al. An architectural overview of the alpha real-time distributed kernel. In Workshop on Micro-Kernels and Other Kernel Architectures, Seattle, WA, Apr. 1992.

....on knowing the sequence of modules that will be traversed by a data item. In addition, it requires a provision for path specific memory allocators. Both requirements can be accommodated easily in the proposed path model. 4 2.4.2. 2 Migrating Distributed Threads Migrating (distributed) threads [19, 45, 37] address the issue of anonymity of processing that often poses problems in modular systems. Typically, when data enters a new module, information on whose behalf the data is being processed is lost. Since, in the path model, all execution is in the context of paths, they can serve the same purpose ....

....management framework. Examples of code, or fast path optimizations include Synthesis [60] Synthetix [85] PathIDs [56] Protocol Accelerators [107] and integrated layer processing [16, 1] Examples in the second category include processor capacity reserves [64] distributed migrating threads [19, 37], and Rialto activities [52] These related works do not attempt to define an explicit and universal path abstraction, but are a source of interesting examples of how paths can be employed. The work on Synthetix is also interesting in that it introduces the notion of quasiinvariants. ....

Raymond K. Clark, E. Douglas Jensen, and Franklin Reynolds. An architectural overview of the Alpha real-time distributed kernel. In Proceedings of the USENIX Workshop on Micro-Kernels and Other Kernel Architectures, pages 127-- 146, Seattle, WA, April 1992. 166

....implies that the thread of control itself is not protected during the call. RPC is used in most other systems; it requires a thread switch and provides protection by separating the caller s and callee s threads. A large body of work has improved the performance of RPC, in particular local RPC [4] [6,14,11] further explored a model of migrating threads where a single thread of control is used in an RPC. However, the emphasis of this work is on improving performance at the expense of fault isolation. The transaction model proposed by VINO [29] offers another solution for protected control transfer ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An Architectural Overview of the Alpha Real-time Distributed Kernel. USENIX Workshop on Micro-kernels and Other Kernel Architectures, Seattle, WA, April 1992. DRAFT 04/15/97 --- DO NOT DISTRIBUTE 19

....was designed from scratch for resource accountability. In contrast to Scout, RT Mach and Rialto, our work aimed at developing a resource accounting mechanism for traditional UNIX systems with minimal disruption to existing APIs and implementations. The migrating threads of Mach [17] and AlphaOS [13], and the shuttles of Spring [19] allow the resource consumption of a thread (or a shuttle) performing a particular independent activity to be charged to the correct resource management entity, even when the thread (or shuttle) moves across protection domains. However, these systems do not ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An Architectural Overview of The Alpha Real-Time Distributed Kernel. In Workshop on Micro-Kernels and Other Kernel Architectures, Apr. 1992.

....is similar to that provided by a number of operating system abstractions developed in the context of multimedia and real time operating systems. These include the processor capacity reserves of Mercer et al. 28] the activities [22] of Rialto, the migrating threads of Mach [17] and AlphaOS [13], and the shuttles of Spring [20] The chief differences are related to the more general nature of resource containers. Processor capacity reserves and activities are real time abstractions and are thus more complex to implement than resource containers. Also, migrating threads, processor ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An architectural overview of the alpha real-time distributed kernel. In Workshop on Micro-Kernels and Other Kernel Architectures, pages 127--146, Seattle WA (USA), Apr. 1992. Usenix.

....systems, services are more concerned with maximizing overall utility (by serving the most important request first) than guaranteeing reserved resources for individual requests. Priority driven services can generally be categorized this way, and are supported in real time kernels such as Alpha [8] and Mach [9] Under overload conditions, lower priority tasks are denied service in favor of more important tasks. In the Rialto operating system [10] a resource planner attempts to dynamically maximize user perceived utility of the entire system. However, the scheme does not adopt the notion of ....

R. Clark, E. Jensen, and F. Reynolds, "An architectural overview of the Alpha real-time distributed kernel," in Proceedings of the USENIX Workshop on Microkernels and other Kernel Architectures, 1992.

....a migrating thread model very similar to ours, although it uses different terminology. Spring s shuttle corresponds to our thread, and their thread corresponds to our activation. Spring addresses the controllability issues but did not have to be concerned with backwards compatibility. Alpha[11] was probably the first system to fully adopt migrating threads. It is oriented to real time constraints, and its migrating thread abstraction is especially important for carrying along scheduling, exceptionhandling, and resource attributes. In both of these systems a thread can migrate across ....

....several advantages which are outlined in this section. The majority of the benefits are linked to use with RPC and are described first. But there are also controllability advantages for threads during all kernel interaction, and these are outlined in section 3.2. In the context of the Alpha OS, [11] also discusses many advantages offered by migrating threads. 3.1 Remote Procedure Call Many of the advantages of migrating threads stem from their use in conjunction with RPC. Migrating threads provide a more appropriate underlying abstraction on which to build RPC interfaces than do static ....

[Article contains additional citation context not shown here]

Raymond K. Clark, E. Douglas Jensen, and Franklin D. Reynolds. An architectural overview of the Alpha real-time distributed kernel. In Proc. of the USENIX Workshop on Micro-kernels and Other Kernel Architectures, pages 127--146, Seattle, WA, April 1992.

....is sometimes used to optimize common path code sequences [24, 23] Specialization, in turn, depends on the existence of invariants that constrain the path through the code that is likely to be executed. ffl The Alpha OS allows threads to migrate across a sequence of protection domains [5]; others have defined similar mechanisms [13, 9] Such mechanisms recognize that tasks often span multiple domains, and so account for resource usage on a path basis rather than a domain basis. The thesis of this paper is that these mechanisms are not isolated optimizations, but rather, that they ....

....ffl improving resource management. Examples of fast path optimizations include Synthesis [19] Synthetix [24] PathIDs [17] Protocol Accelerators [29] and integrated layer processing [4, 1] Examples in the second category include processor capacity reserves [21] distributed migrating threads [5, 9], and Rialto activities [16] Because space does not permit us to contrast all of this work in detail, we simply point out that the path abstraction as presented in this paper is an attempt at unifying these various ideas. In particular, the proposed abstraction allows us to reason about both the ....

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An architectural overview of the Alpha real-time distributed kernel. In 1993 Winter USENIX Conf., pages 127--146, Apr. 1993.

....transparently propagates its timeliness properties (and perhaps also resource ownership, transactional context, security attributes, etc. when its execution point transits object (and perhaps node) boundaries. The distributed thread model appeared first in the Alpha distributed realtime OS kernel [4,11] and was subsequently incorporated in the kernels of the MK7.3 distributed real time OS [26] it is the basis of the programming model for OMG s recent Real Time CORBA 2.0 (Dynamic Scheduling) specification [14] Realistic non trivial experimental distributed real time computing systems were ....

R.K. Clark, E.D. Jensen and F.D. Reynolds, An Architectural Overview of the Alpha Real-Time Distributed Kernel, USENIX Workshop on Microkernels and other Kernel Architectures, pp 200- 208, 1993.

.... scheduling policy and the threads it schedules co exist with and interact with other parts of the OS and application, including synchronizers, preemptions, and priority modifications (e.g. priority inheritances and priority depressions) MK7 threads are distributed (or migrating) threads[2][3] whichmove among the processes (i.e. MK tasks) of a distributed system by executing RPCs while carrying an environment that includes information like the thread s scheduling parameters, identity, and security credentials. MK7 provides several standard interfaces including a highly ....

Clark, R.K., Jensen, E.D., Reynolds, F.D.: An Architectural Overview of the Alpha Real-Time Distributed Kernel. Proc. of the USENIX Workshop on Micro-kernels and Other Kernel Architectures (1992) 127-146

No context found.

R.K. Clark, E.D. Jensen and F.D. Reynolds, An Architectural Overview of the Alpha Real-Time Distributed Kernel, USENIX Workshop on Microkernels and other Kernel Architectures, pp 200208, 1993.

....of analysis and implementation. We are developing and exercising technology that directly addresses how to build and operate computer systems that can adapt to changing circumstances. Key components of our approach include the use of a multi dimensional QoS model [LAW 98] distributed threads [CLA 92] and value based resource management [JEN 85] The operating system being used is the Open Group Research Institute s MK 7 [WEL 94] To exercise this technology, we are developing a prototype of an adaptive AWACS surveillance tracking application. This application adapts its behavior, by ....

R.K. Clark, R.K, Jensen, E.D., Reynolds, F.D., "An Architectural Overview of the Alpha Real-Time Distributed Kernel," Proceedings of the USENIX Workshop on Micro-kernels and Other Kernel Architectures, pp. 127-146, Seattle, WA, April 1992.

....88 Adaptive Distributed Thread Integrity 5.1 Alpha Thread Maintenance and Repair Distributed threads are a programming model abstraction that is used to represent distributed computations. To be concrete, we will consider distributed threads in the context of the Alpha programming model [38, 11], which was designed for real time supervisory control. Here, the programming model is provided directly by the Alpha distributed real time operating system kernel, and distributed threads represent real time distributed computations. When we refer to Alpha, we will usually be referring to the ....

Raymond K. Clark, E. Douglas Jensen, and Franklin D. Reynolds. An Architectural Overview of the Alpha Real-Time Distributed Kernel. In Proceedings of the USENIX Workshop on Microkernels and Other Kernel Architectures, Seattle, WA, April 1992.

No context found.

R. Clark, D.E. Jensen, and F.D Reynolds, An Architectural Overview of the Alpha Realtime Distributed Kernel, In Proceedings of the USENIX Workshop, In 1993.

No context found.

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An Architectural Overview of the Alpha Real-Time Distributed Kernel. In Proceedings of the USENIX Workshop on Microkernels and Other Kernel Architectures, April 1992.

No context found.

R.K. Clark, E.D. Jensen, and F.D. Reynolds, "An Architectural Overview of the Alpha Real-Time Distributed Kernel," Proceedings of the USENIX Workshop on Microkernel and Other Kernel Architectures, Seattle, April 1992.

No context found.

R. K. Clark, E. D. Jensen, and F. D. Reynolds. An Architectural Overview of the Alpha Real-Time Distributed Kernel. In Proceedings of the USENIX Workshop on Microkernels and Other Kernel Architectures, April 1992.

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