International Business Machines Corporation

Susan Patricia Paice

Donald W Schmidt

Elliot Karl Kolodner

Alan Michael Webb

Martin John Trotter

Ethan Lewis

Elizabeth Anne Hutchison

Donna N Dillenberger

Samuel David Borman

Robert F Berry

Rajesh Ramkrishna Bordawekar

Malolan Chetlur

Robert Francis Berry

David Francis Bacon

Tian Cheng Liu

David John Vines

Zygmunt A Lozinski

Andrew John Schofield

Qi Ming Teng

Stuart J Hayton

Jens Diedrichsen

Edmund James Whittaker West

Donna Ngar Ting Dillenberger

A processor-implemented method manages a data cache for a computer system. One or more processors maintain a cache for one or more data resources requested by a client process. In response to a request from the client process for a cached data resource, the processors send a first request to a server process for the data resource, and a second request to the server process for an indication of whether or not the data resource has been modified since the data source was most recently cached. Responses to the first or second requests from the server process are compared with the cached data resources so as to identify whether the response to the second request is inaccurate. One or more processors disable caching for the data resource in response to a predetermined number of inaccurate second requests associated with the data resource being identified.

Managing a data cache for a computer system

A method and system for increasing a system's performance and achieving improved memory utilization by approximating the memory sizes that will be required for data objects that can be deserialized and constructed in a memory cache. The method and system may use accurate calculations or measurements of similar objects to calibrate the approximate memory sizes.

Method and system for approximating object sizes in an object-oriented system

In a virtual machine environment, a method and apparatus for the use of multiple heaps to retain persistent data and transient data wherein the multiple heaps enables a single virtual machine to be easily resettable, thus avoiding the need to terminate and start a new Virtual Machine as well as enabling a single virtual machine to retain data and objects across multiple applications, thus avoiding the computing resource overhead of relinking, reloading, reverifying, and recompiling classes. The memory hierarchy includes a System Heap, a Middleware Heap and a Transient Heap. The use of three heaps enables garbage collection to be selectively targeted to one heap at a time in between applications, thus avoiding this overhead during the life of an application.

Serially, reusable virtual machine

A computer system includes a virtual machine supporting an object-oriented environment, in which programs to run on the virtual machine are formed from classes including methods. Applications run on the virtual machine. After a first such application has run on the virtual machine, a reset is performed on the virtual machine. The reset calls a first method in a class to perform a tidy-up operation which effectively puts the virtual machine back into a clean state. Subsequently a second application can be started on the virtual machine. As part of this start-up, the system calls a second method in the class to perform a reinitialization of the class.

Virtual machine with reset operation

A computer system provides an object-based virtual machine environment for running successive applications. The computer system includes storage, at least a portion of which is logically divided into two or more heaps in which objects can be stored. A first heap is reset between successive applications, and a second heap persists from one application to the next. A card table is provided which comprises multiple cards, each corresponding to a region of said storage. Each card in the card table is set to null when the first heap is reset between successive applications. A card is marked whenever an object in its corresponding storage region is created or updated. It is then possible to detect potential references from the second heap to the first heap at reset by scanning the cards in the card table corresponding to the second heap, and detecting any cards which have been marked. The system further identifies any objects on the first heap which have a finalization method. The finalization methods of any such identified objects are then run on the main thread prior to reset of the first heap.

Computer system with heap reset for performing generational garbage collection implemented by card-marking between successive applications

A computer system provides an object-based environment. The computer system includes storage. A contiguous linear portion of the storage is logically divided into first and second heaps located at opposite ends of the linear portion of storage. Any gap between the two heaps represents an unallocated region of storage. The system permits references from objects on the first heap to objects on the second heap and vice versa. A garbage collector operates across both heaps to remove objects that are no longer live. Means are provided for expanding the first and second heaps into the unallocated region. The first heap is expanded according to a first expansion policy, and the second heap is expanded according to a second expansion policy.

Computer system with two heaps in contiguous storage

A computer system provides an object-based environment and includes storage. At least a portion of the storage is logically divided into two or more heaps in which objects can be stored. Each heap is subdivided into slices of memory. The system includes a two-level lookup structure for determining whether a given storage address corresponds to a particular heap. The lookup substructure involves a first level having one or more lookup substructures, each corresponding to a unit of memory representing a predetermined number of slices. The substructure indicates for each of these slices the particular heap, if any, that the slice belongs to. The two-level lookup structure further involves a second level for determining for a given memory address the first level lookup substructure that includes the slice containing that address.

Computer system with multiple heaps

A system includes first and second virtual machines (VM) representing a master and a client respectively. A class file is loading by the first virtual machine into shared memory. The class file includes sharable and non-sharable data, and also references sharable and non-sharable data (including code). When it is desired to load the same class into the second virtual machine, the class is first located within said shared memory. A mirror copy of the class is then formed within private memory in the second virtual machine. Finally the non-sharable data of the class is completed in the private memory. Thereafter, an application in the second virtual machine can utilise the class on the basis of the sharable data in the shared memory of the first virtual machine, and the non-sharable data from the private memory of the second virtual machine.

Class sharing between multiple virtual machines

Extended business transactions are processed in a client-server system in a manner which allows processing initiated by a client to be interrupted before the transaction is complete and later resumed by the same or another client from the point of interruption. This is achieved by storing state information indicative of the progress of the transaction in a repository in association with an end-user identifier. When the end-user communicates his identifier via a client to the server for a second time, processing of the transaction can be resumed on the basis of the stored state.

Growing community of inventors

Eastleigh, United Kingdom

Susan Patricia Paice