UNIX internals: the new frontiers
UNIX internals: the new frontiers
Evaluation of design choices for gang scheduling using distributed hierarchical control
Journal of Parallel and Distributed Computing
Effective distributed scheduling of parallel workloads
Proceedings of the 1996 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Adaptive algorithms for managing a distributed data processing workload
IBM Systems Journal
An infrastructure for efficient parallel job execution in Terascale computing environments
SC '98 Proceedings of the 1998 ACM/IEEE conference on Supercomputing
Demand-Based Coscheduling of Parallel Jobs on Multiprogrammed Multiprocessors
IPPS '95 Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing
The ANL/IBM SP Scheduling System
IPPS '95 Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing
The EASY - LoadLeveler API Project
IPPS '96 Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing
Improved Utilization and Responsiveness with Gang Scheduling
IPPS '97 Proceedings of the Job Scheduling Strategies for Parallel Processing
Expanding Symmetric Multiprocessor Capability Through Gang Scheduling
IPPS/SPDP '98 Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing
Implementing the Combination of Time Sharing and Space Sharing on AP/Linux
IPPS/SPDP '98 Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing
Improving First-Come-First-Serve Job Scheduling by Gang Scheduling
IPPS/SPDP '98 Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing
Dynamic Coscheduling on Workstation Clusters
IPPS/SPDP '98 Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing
Gang scheduling for highly efficient, distributed multiprocessor systems
FRONTIERS '96 Proceedings of the 6th Symposium on the Frontiers of Massively Parallel Computation
Extensible Resource Management For Cluster Computing
ICDCS '97 Proceedings of the 17th International Conference on Distributed Computing Systems (ICDCS '97)
Hi-index | 0.00 |
Job management subsystems in parallel environments have to address two important issues: (i) how to associate processes present in the system to the tasks of parallel jobs, and (ii) how to control execution of these tasks. The standard UNIX mechanism for job control, process groups, is not appropriate for this purpose as processes can escape their original groups and start new ones. We introduce the concept of genealogy, in which a process is identified by the genetic footprint it inherits from its parent. With this concept, tasks are defined by sets of processes with a common ancestor. Process tracking is the mechanism by which we implement the genealogy concept in the IBM AIX operating system. No changes to the kernel are necessary and individual process control is achieved through standard UNIX signaling methods. Performance evaluation, on both uniprocessor and multiprocessor systems, demonstrate the efficacy of job control through process tracking. Process tracking has been incorporated in a research prototype gang-scheduling system for the IBM RS/6000 SP.