Dearle, A., et al "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, Vol 7(3), Summer 1994  @ NUS  Home/Search   Document Details and Download   Summary   Related Articles   Check

....of kernel memory and can access this part directly using mappings in its address space. The checkpoint server is also allowed to revoke certain privileges from the kernel, e.g. to implement copy on write for kernel metadata. The Charm microkernel [12] which is derived from the Grasshopper system [11], exposes its metadata to user level in order to support applications in implementing their own persistence policy. However, applications are not allowed to create or modify metadata; security related metadata is not exposed. 2.3 Summary In short, previous approaches to kernel memory management ....

Alan Dearle, Rex di Bona, James Farrow, Frans Hensken, Anders Lindstrm, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994.

....are included within the checkpoint. Fluke [14] is a kernel that exports user visible, partly pickled kernel objects to the user. This way a user level checkpointer pickles the remaining parts of the kernel objects and save these along with the memory image of the tasks. Grasshopper [16] is a larger system, because it is not based upon a kernel and uses abstractions specially designed with persistence in mind. These kernel abstractions are used to implement persistence on kernel level. Based on the experience with Grasshopper its designers later created a kernel based ....

Alan Dearle, Rex di Bona, James Farrow, Frans Hensken, Anders Lindstrom, John Rosenberg and Francis Vaughan, Grasshopper: an orthogonally persistent operating system, Computing Systems, 7(3):289-312, Summer 1994.

....of using a pointer. So called Single Address Space Operating Systems (SASOS) such as Monads [15] Opal [8] This work was supported in part by the Esprit Project PerDiS N 22533. E. Jul (Ed. ECOOP 98, LNCS 1445, pp. 234 259, 1998. c Springer Verlag Berlin Heidelberg 1998 or Grasshopper [10] extend this simple model elegantly to distribution and persistence. In a SASOS, an object is mapped at the same address in every process ever accessing it, ensuring that pointers remain valid across address spaces and time. It uses DSM techniques to ensure consistency of distributed replicas, and ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994. 235

....uservisible, partly pickled, kernel objects. The checkpointer pickles the remaining parts of the objects and saves them to stable storage together with the memoryimages of the tasks. Most transparently persistent operating systems are based upon kernels. An exception to this rule is Grasshopper [2]. Grasshopper hopes to achieve persistence through the use of some specially designed kernel abstractions. Based on the experiences learned from Grasshopper, the designers have later created a kernel based operating system, Charm [3] aiming at supporting persistent applications. With Charm, no ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: an orthogonally persistent operating system. Computing Systems, 7(3):289-312, Summer 1994.

....are retained as long as you keep paying the rent. The Clouds operating system [56] provided persistent objects and persistent threads on a distributed network of standard computers. Similarly, the KeyKos [84] and Eros [190] operating systems offer transparent persistence for objects. Grasshopper [71] is a general purpose persistent platform. The challenge of getting applications built on an unfamiliar operating system, inhibited the use of persistent operating systems as vehicle with which to test the OP hypothesis. It seems probable that operating systems supporting persistence will become ....

....is why we have kept them out of the OPJ specification until we are persuaded that they are really necessary. 12.2.2 Normal Execution Performance Ideally no cost would be paid for persistence once the working set of objects needed by an application have been faulted into main memory. Some systems [191, 207, 166, 71, 133, 101] approach this goal by utilising the memory management hardware to detect both non resident objects, and mutated objects for checkpoint purposes. It is currently the case that the cost of servicing the faults, when they do occur, can be very high with existing commercial operating systems, such ....

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrom, J. Rosenberg, and F. Vaughan. Grasshopper: An Orthogonally Persistent Operating System. Computing Systems, 7(3):289--312, 1994.

....systems, shared object stores, fault tolerant distributed systems and approaches based on a special purpose hardware architecture. In the area of general persistent programming PS Algol[16] E[29] Napier88 [11] Texas[33] may be mentioned. In the operating systems area systems like Grasshopper[12], Choices[6] Chorus Cool2 [1] and Clouds[10] support persistent programming ( 13] gives an analysis on operating system support for 16 persistence) whereas Monads[30] provides a single level persistent store based on special hardware features. Robust distributed programming is addressed by ....

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrom, J. Rosenberg, and F. Vaughan. Grasshopper: An Orthogonally Persistent Operating System. Computing Systems, 7(3), 1994.

....on improving efficiency in the model, but rather to show that persistent stores and the persistent programming model has advantages compared to relational databases or remote object invocation in CORBA. Examples of distributed persistent stores and operating systems are PerDis [5] and Grasshopper [6]. 6. Further work The implementation presented in this article can be further improved. At the moment all writes are immediately performed, even if they are to a locked object. This situation can be improved by not having a write through cache in the clients. Updates can be performed globally at ....

Alan Dearle, Francis Vaughan, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrm and John Rosenberg, Grasshopper: An orthogonally persistent operating system, Tech.Report GH10, Dept. of Computer Science University of Adelaide, URL:http://www.gh.cs.su.oz.au/Grasshopper/Papers/GH10/gh10.html, 1994

....also clear. DSMs have been most successful in supporting multi threaded parallel programs, whereas PerDiS facilitates sharing between di erent programs. Many of the ideas in PerDiS are direct descendents of the so called SingleAddress Space Operating Systems (SASOS) such as Opal [11] Grasshopper [14], or EOS [22] The main di erence with PerDiS lies in the integration of persistent object systems and object oriented database concepts, e.g. transactions, persistence by reachability and garbage collection. This is why our system uses transactions and incorporates a security architecture. We ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289-312, 1994.

.... offer persistence as a first class feature (orthogonal persistence [25] can make some applications much easier to implement [17] Unfortunately, orthogonal persistence maps clumsily to a conventional file system, so sometimes orthogonal persistence is provided directly via the operating system [15, 21, 9]. These examples include one solution that works by rewriting the binary code of methods and others that work by making persistent or replicated the entire environment that objects run inside. With a general partial evaluator, we would expect to see similar approaches, but defined in a more ....

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrom, J. Rosenberg, and F. Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, Summer 1994.

....well. To achieve Department of Computing Science, Chalmers University of Technology and Gteborg University E mail: bengtj cs.chalmers.se, WWW: http: www.cs.chalmers.se bengtj this, an efficient caching scheme must be used. To date, the systems we are aware of, are naive in this respect. [5,6] This article describes how a distributed persistent store may take advantage of an efficient caching scheme. The system we describe uses a client server model to implement the store. One or more clients communicate with servers containing the persistent data. See fig.1) Clients may run on the ....

....to show that persistent stores and the persistent programming model has advantages compared to relational databases or remote object invocation in CORBA. For instance, the distributed persistent store PerDis [5] uses sequential consistency in its caches. The persistent operating system Grasshopper [6] appears not to address this issue at all. However, it is unlikely that Grasshopper does not use any kind of caching. 7. Conclusions Relaxed consistency models are much used in multiprocessor implementations. They lead to decreased memory latency and makes it possible to take advantage of ....

Alan Dearle, Francis Vaughan, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrm and John Rosenberg, Grasshopper: An orthogonally persistent operating system, Tech.Report GH10, Dept. of Computer Science University of Adelaide, URL:http://www.gh.cs.su.oz.au/Grasshopper/Papers/GH10/gh10.html, 1994

....data (DSM) Distinct node namespaces for code make task migration difficult. 2.3. SASOSs with and without changed protection hardware support Most contemporary SASOSs have been designed to run on stock RISC 64 bit processors. These include Opal [1] 3] Mungi [4] Arius [5] and Grasshopper [6]. This means additional software communication protocols are required for protection. Typically capabilities are used requiring operations in one or more additional namespaces. SASOSs with protection and memory management hardware targeted at single address space operation including AS 400 [7] ....

....store and a flat 60 bit paged single virtual address. The MONADS hardware predates modern 64 bit RISC technology. Researchers at the University of Sydney are currently working on the Grasshopper orthogonal persistent object operating system, which is targeted to run on stock 64 bit RISC processors [6]. 8. Summary The SASOS concept and in particular Sombrero provide a low level substrate that substantially reduces the problems to clustered computing associated with process oriented architectures. In particular communication between cooperating nodes in a cluster becomes transparent allowing ....

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindstrm, A., Rosenberg, J. and Vaughn, F., "Grasshopper: An Orthogonally Persistent Operating System", Computing Systems, 7(3), pp. 289-312, Summer 1994.

....can be supported without redesign and reimplementation of the programming language if one is prepared instead to layer support for persistence into the operating system. Several experimental projects have taken this approach: support for persistence is targeted explicitly in Grasshopper [29, 71] and Mungi [34, 37] but the rudiments are there in other experimental operating systems such as Opal [24, 25] among others. Our interest here focuses on efficient support for orthogonal persistence on stock operating systems. 4 Mostly copying garbage collection Mostly copying garbage ....

DEARLE,A.,DI BONA,R.,FARROW,J.,HENSKENS,F., LINDSTR OM,A.,ROSENBERG,J.,AND VAUGHAN,F. Grasshopper: An orthogonally persistent operating system. Computer Systems 7, 3 (Summer 1994), 289--312.

....in performance makes a switch to incompatible hardware worthwhile [6] and to a new design paradigm. There have been a number of SASOS designs with and without additional protection hardware support. Those without additional hardware include Opal [2] 7] Mungi [8] Arius [9] and Grasshopper [10]. Those with modified protection hardware include AS 400 [11] Monads [12] and Sombrero [13] See [14] 15] for comparisons between the features of hardware supported SASOSs and those built on stock RISC processors and between Sombrero and other hardware supported SASOSs. 3. Sombrero design ....

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindstrm, A., Rosenberg, J. and Vaughn, F., "Grasshopper: An Orthogonally Persistent Operating System", Computing Systems, 7(3), pp. 289-312, Summer 1994.

....(1) 39 63 53 72 essentially of a large scale DSM that is persistently backed to disk and garbage collected. 1.1 Motivation In a centralized program, sharing consists simply of using a pointer. So called Single Address Space Operating Systems (SASOS) such as Monads [19] Opal [9] or Grasshopper [11] extend this simple model elegantly to distribution and persistence. In a SASOS, an object is mapped at the same address in every process ever accessing it, ensuring that pointers remain valid across address spaces and time. DSM techniques ensure consistency of distributed replicas, while ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994.

....to disk and garbage collected. 1.1 Motivation In a centralized program, sharing consists simply of using a pointer. So called Single Address Space Operating Systems (SASOS) such as Monads [15] Opal This work was supported in part by the Esprit Project PerDiS N o 22533. 8] or Grasshopper [10] extend this simple model elegantly to distribution and persistence. In a SASOS, an object is mapped at the same address in every process ever accessing it, ensuring that pointers remain valid across address spaces and time. It uses DSM techniques to ensure consistency of distributed replicas, and ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994.

....of cache coherence protocols. This paper presents a protocol customization system (PCS) of memory mapped distributed file caches, which 1 Pointer conversion is implemented more efficiently than in the past because of the evolution of lazy conversion techniques called pointer swizzling [19, 5]. Nevertheless it is one of the major bottlenecks of persistent systems. enables distributed cooperative applications to optimize the behavior of file caches in an application specific manner. PCS is integrated in the distributed cache management layer of a distributed file system, and allows ....

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstr om, J. Rosenberg, and F. Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, Summer 1994.

....also clear. DSMs have been most successful in supporting multi threaded parallel programs, whereas PerDiS facilitates sharing between dioeerent programs. Many of the ideas in PerDiS are direct descendents of the so called Single Address Space Operating Systems (SASOS) such as Opal [9] Grasshopper [12], or EOS [21] The main dioeerence with PerDiS lies in the integration of persistent object systems and object oriented database concepts, e.g. transactions, persistence by reachability and garbage collection. This is why our system uses transactions and incorporates a security architecture. We ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstr#m, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289312, 1994.

....for uninitialised memory. Fig. 3 shows an example of an Unmap operation. ABC PM: VM: ABC ABC from to ABC PM: ABC VM: from to Figure 3: Address space before (left) and after (right) execution of an Unmap(from,length) system call. 2. 5 Recursive address spaces Systems like Grasshopper [9] and L4 [10] use similar mapping operations, but use them in a hierarchical fashion to recursively construct address spaces. This does not fit the SASOS model, as there can only be one address space, so mappings can only operate between different parts of the same address space. However, pagers ....

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrom, and F. Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994.

.... instead of distributing virtual memory on top of centralized physical memory management. Advice provides a flexible framework allowing user level applications to adapt distributed virtual and shared memory [8, 14] to their requirements. Approaches like those found in some persistent systems [10] and software based DSM systems [16] can be implemented using Advice. We can also take advantage of other flexible memory management models [13] while each application will use and pay only for the distributed virtual memory services it needs. The next section is a general overview of the Advice ....

Alan Dearle. Grasshopper: An orthogonally persistent operating system. Computing Systems, 3(7):289--312, 1994.

....is constructed. This may be a persistent operating system (on top of which, all languages achieve persistence automatically, and with which all data has consistent persistent behaviour) or even new hardware architectures with associated operating systems. Typical of the former are: Grasshopper [Dearle et al. 1994], EOS [Gruber, 1992, Dayn s Gruber, 1994] also a FIDE 2 product) and of the latter are: Rosenberg s proposals [Rosenberg, 1990] and DAIS [Russell et al. 1994] The pay off from investing in good persistent technology is only apparent when a large application is used over a long period. Hence ....

....over the locality of data, that is, the programmer does not need to be aware of whether it is on disk or in RAM. One approach to distributed orthogonally persistent systems is to continue this location transparency [Wai, 1989, Koch, et al. 1990, Gruber et al. 1992, Dayn s and Gruber, 1994, Dearle, et al. 1994]. This abstraction may eventually satisfy those that use distribution to achieve reliability and performance through distributing computation and migrating computation or data, but it depends on the discovery of adequate algorithms for 49 automating data and program movement. Although this ....

[Article contains additional citation context not shown here]

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindström, A., Rosenberg, J. & Vaughan, F., 1994. Grasshopper: An Orthogonally Persistent Operating System. Computer Systems 7, 3 pp 289-312.

....these problems while retaining a lean design. Instead of distributing virtual memory on top of centralized physical memory management, Advice approach is to implement distributed virtual memory on top of distributed physical memory management. Approaches like those found in persistent systems [3] and other (software based) DSM systems [7] can be also implemented using the Advice approach. Advantages of other flexible memory management models [5] can thus be found in Advice. Now that the idea behind is stated, the architecture will be introduced. 3 Overall Design In Advice, each node has ....

Alan Dearle et al. Grasshopper: An orthogonally persistent operating system. Computing Systems, 3(7):289--312, 1994.

.... problem of this approach is that we cannot expect a single cache coherent 1 Pointer conversion is one of the major bottlenecks of persistent systems, even though we can implement it more efficiently than it once was because of the evolution of lazy conversion techniques called pointer swizzling [33, 32, 9]. protocol to serve a wide range of cooperative applications efficiently. A number of DSM coherence protocols have been proposed so far, yet no single protocol is known to be the best. Distributed cooperative applications, in particular, have much opportunity for application specific ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, Summer 1994.

....include compiler support and programming environments, including specific debuggers [57] A compelling reason (in our opinion) whyDSMs are not more widespread is the lack of persistence and poor support for sharing complex data types. An extension of the DSM concept is the Shared Address Space [11, 14, 39, 42], which allows programs to share data by using ordinary pointers. Pointer swizzling [32, 53] provides independence from any particular addressing convention, while also enabling sharing of pointers. This extremely simple addressing model is the basis of our PDS concept. INRIA SOR has improved on ....

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstr om, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994.

....kernel drivers only support those operations that are required for the controllers. The design of the structure arose out of the desire to produce a simple, generic, driver for the Grasshopper project. The Grasshopper project is designing a persistent operating system for experimental purposes [5]. As such, one of the project goals is to design a system that allows experimentation with a variety of different high level systems. The persistent nature of Grasshopper is not crucial in the design of the disk interface, and we feel the interface is equally suited to non persistent operating ....

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindstroaom,A., Rosenberg, J. and Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System", Technical Report,Departments of Computer Science, Universities of Adelaide and Sydney, 1993.

....of purposes in our model. First, they provide a means for data storage. The data in a container is persistent, meaning that its lifetime is orthogonal to the lifetime of the creating locus and that it survives system shutdowns. The details of container data management are further discussed in [7] . Second, containers provide address spaces for loci which execute within them. A single locus may move between containers by the invocation mechanism explained below. The container in which a locus currently executes is called its host container. During execution, all virtual addresses issued by ....

....of the scheme by showing how shared libraries, the Unix model and a single address space environment may be implemented above the new model. Although space has not permitted us to discuss persistence, the model also includes mechanisms to support user level control over the lifetime of data [7] . The model described has been implemented as part of the Grasshopper operating system project at the Universities of Sydney and Adelaide. The first implementation is on the DEC Alpha platform and is operational in a limited form. A prototype of the Unix server also exists. Initial timing ....

Dearle, A., di Bona, R., Farrow, J. M., Henskens, F. A., Lindstrm, A., Rosenberg, J. and Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, vol 7, 3, pp. 289-312, 1994.

....few years and describes our plans to produce a new persistent micro kernel. Keywords: operating systems, micro kernels, exo kernels, persistence. 1. Introduction During the past five years we have played a key role in the design and construction of the Grasshopper persistent operating system [17, 41]. As the implementation matured, we began to gain valuable experience in the use of the basic abstractions by porting various application systems [15, 18] that could benefit from orthogonal persistence. At the same time, development work was proceeding on the operating system kernel to experiment ....

....in this manner can only have an adverse effect on performance. There are numerous examples of operating systems that support persistence in one form or another. Examples include Multics [5, 13, 37] MONADS [26, 39, 40] Eumul L3 [28] Clouds [14] Choices [9] KeyKOS [8, 24] and Grasshopper [17, 41]. For the past five years, we have been involved in the design and implementation of the Grasshopper system, an overview of which is presented below. 3. Grasshopper Grasshopper is an operating system explicitly designed to support orthogonal persistence. To this end, it provides three ....

[Article contains additional citation context not shown here]

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrm, J. Rosenberg, and F. Vaughan, "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, Summer, pp. 289-312, 1994.

....to be recorded, and later browsed and or retrieved by the user. This architecture permits more sophisticated browsing and program development tools to be constructed. The operating system used in implementing the architecture described in this paper is the persistent operating system Grasshopper [6]. Section 2 gives an outline of this operating system and introduces the abstractions pertinent to the rest of the paper, as well as the IDL model used to represent types in the architecture. Section 3 describes the architecture, Section 4 examines the G shell, a tool which is built using the ....

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindstrm, A., Rosenberg, J. and Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, vol Summer, pp. 289-312, 1994.

....regions within their address spaces. These are typically used to provide access to per locus data structures such as stacks. From an external view of the Grasshopper system, it appears as if all data is stored within containers. In reality however, the actual data is maintained by managers [22]. A manager is a distinguished container holding code and data to support the transparent movement of data between primary and secondary storage. They are the only component of Grasshopper in which the distinction between long and short lived data is apparent. To support the implementation of ....

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrm, J. Rosenberg, and F. Vaughan, "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, Summer, pp. 289312, 1994.

....and the programmer is forced to wrestle with the operating system in order to achieve the desired result. A persistent operating system provides a natural and elegant solution, whilst maintaining efficiency. This paper describes the approach to persistence and resilience taken in Grasshopper [9], a persistent operating system being developed at the Universities of Stirling and Sydney. We first describe the various approaches to data management and demonstrate that the approach used in persistent systems removes the need for ad hoc techniques. This is followed by a discussion of the ....

....Choices [4] takes a different approach in which the basic kernel abstraction of storage is the memory object. Inheritance is then used to special these into both file objects and persistent objects. Again, persistence is not ubiquitous and processes are transient. Grasshopper Grasshopper [9] is an example of a persistent operating system. In this section we describe the three basic abstractions provided by Grasshopper. The abstraction over storage is the container and the abstraction over execution is the locus (plural loci) The container in which a locus is currently executing is ....

Dearle, A., di Bona, R. M., Farrow, J. M., Henskens, F. A., Lindstrm, A. G., Rosenberg, J. and Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, vol 7, 3, pp. 289-312, 1994.

....address the problem with respect to network transmission. The final approach to providing persistence is to provide it at the virtual address space level; this approach is followed in the design and implementation of Grasshopper and has many desirable properties which we have described elsewhere [5]. One benefit of this approach is the ability to make all data in the address space persistent, including the state of threads and the stacks supporting them. Curiously, this does not assist in the transmission of state to another machine since data must be in an architecturally neutral format to ....

Dearle, A., et al., Grasshopper: An Orthogonally Persistent Operating System. Computer Systems, 1994. Summer: p. 289-312.

....in order to provide a more appropriate environment [9,17,33] These groups have encountered difficulties because of the cost of building hardware using the latest technology and the problems associated with making the results of the research available to other research groups. In Grasshopper [12,13] we have adopted a third approach which is to develop a new operating system on a conventional workstation platform. Some other research groups have also taken this route [5,6,10] We see the advantages of this approach as: workstations are cheap and readily available, their performance is ....

....manager. Managers are responsible for maintaining a consistent and recoverable stable copy of the data represented by the container. The use of managers, which is vital to the removal of the distinction between persistent and volatile storage, is beyond the scope of this paper and is discussed in [12]. 2.1 Containers Containers are the only storage abstraction provided by Grasshopper; they are persistent entities which replace both address spaces and file systems. In most operating systems, the notion of a virtual address space is associated with an ephemeral entity, a process, which ....

Dearle, A., di Bona, R., Farrow, J. M., Henskens, F. A., Lindstrm, A., Rosenberg, J. and Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System ", Computer Systems (to appear), 1994.

....data, 2. some naming mechanisms must be provided for finding the files, and 3. some protection mechanism must be provided to prevent unauthorised access to files and to prevent unauthorised modification. In Grasshopper the first of these is satisfied by a single abstraction known as a container [11]. Containers are the only storage abstraction provided by Grasshopper and are persistent entities which fulfil the roles traditionally served by both address spaces and files. In most operating systems, the notion of a virtual address space is associated with an ephemeral entity, a process, which ....

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindström, A., Rosenberg, J. and Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, vol Summer, pp. 289-312, 1994.

....4 describes various strategies which can be employed to improve recovery time to adapt to the needs of the system being developed. Finally, Section 5 concludes the paper. 2 2. Grasshopper Grasshopper is an operating system specifically designed to support the development of persistent systems [1]. It provides three basic abstractions, containers, loci and capabilities, all of which are inherently persistent. Containers are the only storage abstraction provided by Grasshopper. They are coarsegrained persistent entities capable of holding very large quantities of data. They have properties ....

....is composed of data stored within a particular container known as its host. Although each locus executes within a separate address space, many loci may share a common host container which allows them to share the stored data. A locus may move to a new host container via the invocation mechanism [1] which effectively changes its host container. For these reasons, containers and loci are completely orthogonal abstractions. The persistence of data stored within containers is handled by managers which are responsible for their stability and resilience. Managers integrate these facilities with ....

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindström, A., Rosenberg, J. and Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, vol Summer, 1994.

No context found.

Dearle, A., et al "Grasshopper: An Orthogonally Persistent Operating System", Computer Systems, Vol 7(3), Summer 1994

No context found.

A. Dearle, et. al. Grasshopper: An Orthogonally Persistent Operating System, Computer Systems, 1994. Summer, pp289-312. 18

No context found.

A. Dearle, et. al. Grasshopper: An Orthogonally Persistent Operating System, Computer Systems, 1994. Summer, pp289-312. 18

No context found.

Dearle, A., Vaughan, F., di Bonda, R., Farrow, J., Henskens, F., Lindstrom, A., Rosenberg, J.: Grasshopper: An orthogonally persistent operating system. Computing Systems 7(3), pp 289--312, 1994

No context found.

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrom, J. Rosenberg, and F. Vaughan. Grasshopper: An Orthogonally Persistent Operating System. Computing Systems, 7(3):289--312, 1994.

No context found.

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, Summer 1994.

No context found.

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrom, J. Rosenberg, and F. Vaughan. Grasshopper: an orthogonally persistent operating system. Computing Systems, 7(3):289-- 312, Summer 1994.

No context found.

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994.

No context found.

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, and Francis Vaughan. Grasshopper: An orthogonally persistent operating system. Computing Systems, 7(3):289--312, 1994.

No context found.

A. Dearle, R. di Bona, J. Farrow, F. Henskens, A. Lindstrom, and J. Rosenberg. Grasshopper: An orthogonally persistent operating system. Computer Systems, 7(3):289--312, 1994.

No context found.

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindstrm, A., Rosenberg, J. & Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System". Computer Systems 7, 3 (1994) pp 289-312. 24

No context found.

Dea94A Dearle, A. & di Bona, R. & Farrow, J. & Henskens, F. & Lindstrom, A. & Rosenberg, J. & Vaughan, F.: Grasshopper: An Orthogonally Persistent Operating System, Computing Systems Vol. 7 No. 3 pp 289-312, ftp://nezz.cs.stir.ac.uk/pub/papers/GH-10.ps.Z, 1994.

No context found.

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally Persistent Operating System. Computer Systems, 7(3):289-312, 1992.

No context found.

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally Persistent Operating System. Computer Systems, 7(3):289-312, 1992. Definition of the Project Requirements - Part I Section A Page 58

No context found.

Alan Dearle, Rex di Bona, James Farrow, Frans Henskens, Anders Lindstrom, John Rosenberg, and Francis Vaughan. Grasshopper: An orthogonally Persistent Operating System. Computer Systems, 7(3):289-312, 1992.

No context found.

A. Dearle, R. Bona, J. Farrow, F. Henskens, A. Lindstrom, J. Rosenberg, F. Vaughan, Grasshopper: An orthogonally persistent operating system, Department of Computer Science, University of Adelaide and Sydney, Austria

No context found.

Dearle, A., di Bona, R., Farrow, J., Henskens, F., Lindström, A., Rosenberg, J. & Vaughan, F. "Grasshopper: An Orthogonally Persistent Operating System". Computer Systems, Summer 1994 pp 289-312.

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.