International Business Machines Corporation

Waiman Chan

Joseph Francis Skovira

David Paul Brelsford

Rama Krishna Govindaraju

Alexander Druyan

Kailash N Marthi

Stephen Craig Hughes

Ravindra R Sure

Margaret M Curry

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

A computing system configured to handle preemption events in an environment having jobs with high and low priorities. The system includes a job queue configured to receive job requests from users, the job queue storing the jobs in an order based on the priority of the jobs, and indicating whether a job is a high priority job or a low priority job. The system also includes a plurality of node clusters, each node cluster including a plurality of nodes and a scheduler coupled to the job queue and to the plurality of node clusters and configured to assign jobs from the job queue to the plurality of node clusters. The scheduler is configured to preempt a first low priority job running in a first node cluster with a high priority job that appears in the job queue after the low priority job has started and, in the event that a second low priority job from the job queue may run on a portion of the plurality of nodes in the first node cluster during a remaining processing time for the high priority job, backfill the second low priority job into the portion of the plurality of nodes and, in the event a second high priority job is received in the job queue and may run on the portion of the plurality of nodes, return the second low priority job to the job queue.

System and method for multi-level preemption scheduling in high performance processing

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 method, grid computing environment, and computer readable medium for managing available resources in a grid computing environment are disclosed. The method includes determining, when a new job is submitted, if a first of the computing clusters has sufficient available resources to run the new job. If the first computing cluster does not have sufficient available resources to run the new job, the method further includes determining if a second of the computing clusters has sufficient available resources to run a first job that is currently running on the first computing cluster. If the second computing cluster has sufficient available resources to run the first job, the first job is migrated to the second computing cluster.

Scale across in a grid computing environment

A facility is presented for scheduling computer jobs of a computing environment having a plurality of computing nodes. The facility includes scheduling N sequential computer jobs of a job queue until achieving a top job party, wherein N&#x2267;2, and the number of jobs in the top job party is dynamically determined by monitoring backfill window coverage of the computing nodes by the N scheduled jobs and automatically closing membership in the top job party once (i) all computing nodes are covered by at least one backfill window, or (ii) a backfill end time of at least one computing node is greater than or equal to a maximum wallclock limit of a set of job classes having the N scheduled jobs. Responsive to determining the top job party, the facility then backfill schedules jobs of the job queue around the top job party for execution.

Scheduling of computer jobs employing dynamically determined top job party

A method is provided for starting a job in a parallel batch job data processing system in which a job is submitted to a scheduler having available, for use, a plurality of resource classes in the data processing system and by also including, within the job, instructions which cause notification of job starting to the scheduler and then, upon this notification, removing the job from all but the first resource class on which it is started.

Job routing to earliest available resources in a parallel job scheduler

A model for computer resources based on type and amount is employed in a consistent manner across several layers in a data processing system as a mechanism for providing users with significantly more control over the allocation, utilization and scheduling of resources for user's software applications running on the target computer system.

Growing community of inventors

Poughkeepsie, NY, United States of America

Waiman Chan