International Business Machines Corporation

David Paul Brelsford

Alexander Druyan

Joseph Francis Skovira

Waiman Chan

Damian Leo Osisek

Kenneth E Plambeck

Joseph M Gdaniec

Enci Zhong

Kailash N Marthi

Richard J Coppinger

Peter H Tallman

Stephen Craig Hughes

Margaret Croft Enichen

Ravindra R Sure

Pamela H Dewey

Paul A Malinowski

Gerald Allen Davison

Roger W Rogers

Melvin M Cutler

Lynn A Czak

Sarah T Hartley

Leslie S Coleman

Daniel D Cerutti

Jean-Louis Lafitte

A system to improve cluster machine processing that may include a plurality of interconnected computers that process data as one if necessary, and at least one other plurality of interconnected computers that process data as one if necessary. The system may also include a central manager to control what data processing is performed on a shared processing job performed by the plurality of interconnected computers and the at least one other plurality of interconnected computers. Each of the plurality of interconnected computers runs parallel jobs scheduled by a local backfill scheduler. In order to schedule a cluster spanning parallel job, the local schedulers cooperate on placement and timing of the cluster spanning job, using existing backfill rules in order not to disturb the local job streams.

System to improve cluster machine processing and associated methods

Start times are determined for jobs to be executed in the future in a multi-cluster environment. The start times are, for instance, the earliest start times in which the jobs may be executed. The start times are computed in logarithmic time, providing processing efficiencies for the multi-cluster environment. Processing efficiencies are further realized by employing parallel processing in determining the start times.

Establishing future start times for jobs to be executed in a multi-cluster environment

The present invention employs a master node for each job to be scheduled and in turn the master node distributes job start information and executable tasks to a plurality of nodes configured in a hierarchical node tree of a multinode job scheduling system. The status of the various tasks executing at the leaf nodes and other nodes of the tree report status back up the same hierarchical tree structure used to start the job, not to a scheduling agent but rather to the master node which has been established by the scheduling agent as the focal point, not only for job starting, but also for the reporting of status information from the leaf and other nodes in the tree.

Dynamic allocation and partitioning of compute nodes in hierarchical job scheduling

A fault tolerant capability for the reliable collection and aggregation of disparate data from multiple processing units. Data of a processing unit is collected and locally aggregated at that processing unit. The locally aggregated data is stored in a local database, and periodically forwarded to a central manager. In response to receiving the data, the central manager determines whether the data should be included in a total aggregate of data for the processing environment.

Fault tolerant facility for the aggregation of data from multiple processing units

External job scheduling capabilities are provided to a local job control system lacking or having insufficient job scheduling capabilities. This is accomplished by encapsulating running of a user job by the local job control system (LJCS) within running of a marker job at a node management system, which is responsive to an external resource scheduler. The technique includes starting a marker job by the resource scheduler external to the local job control system, wherein the marker job corresponds to a user job to be run by the LJCS; responsive to starting of the marker job, starting the user job by the LJCS; and upon termination of the user job, ending the marker job started by the resource scheduler, wherein starting of the user job by the LJCS is responsive to the resource scheduler's starting of the marker job and wherein ending of the marker job is responsive to termination of the user job.

External job scheduling within a distributed processing system having a local job control system

A method and apparatus for providing common access to data spaces by a plurality of virtual machine guests emulated on a host computer system. A token received from a given one of the virtual machine guests is used to identify a particular host data space. The guest also supplies an offset. The offset and the identified host data space are then used to derive a host absolute address representative of a data location in the host data space.

Apparatus and method for providing private and shared access to host

The invention disclosed and claimed herein provides a method for a virtual machine, which maps to the V=R region of a host machine's address space, to resume program execution successfully when the host operating system terminates and subsequently restarts successfully after the occurrence of a system incident. The system incident brings the computer system down, but with a reasonable chance that the system will be able to bounce. A bounce occurs when the host operating system nucleus is reinitialized or refreshed. The virtual machine will be allowed to survive the system incident as long as its integrity can be maintained, i.e. as long as its status and in-progress work can be preserved. The status and in-progress work of the virtual machine is preserved in a reserved region of host storage in such a way that the time to save and restore virtual machine status is reduced, I/O that might cause further system incidents are avoided and the complexity of restoration of the virtual machine during a system incident is reduced.

Recovery of guest virtual machines after failure of a host real machine

Growing community of inventors

Hyde Park, NY, United States of America

David Paul Brelsford