File job_scheduler.h¶
↰ Parent directory (job_management
)
Job scheduler detection and resource management for HPC environments.
Definition (job_management/job_scheduler.h
)¶
Detailed Description¶
Le Nhan Pham
2025
This header provides comprehensive job scheduler detection and resource management for High Performance Computing (HPC) environments. It automatically detects the active job scheduler, extracts resource allocation information, and provides safe resource usage calculations that respect cluster limits.
Schedulers
The system supports detection and resource extraction for:
SLURM (Simple Linux Utility for Resource Management)
PBS/Torque (Portable Batch System)
SGE/OGS (Sun Grid Engine / Open Grid Scheduler)
LSF (IBM Load Sharing Facility)
Generic cluster detection for unknown schedulers
Management
Key features include:
Automatic detection of allocated CPUs and memory
Safe resource limit calculations to prevent oversubscription
Integration with application threading and memory management
Environment variable parsing for all major schedulers
Validation and safety checks for resource requests
Integration
The job scheduler system integrates with:
Multi-threading systems for safe CPU allocation
Memory management systems for memory limit enforcement
Configuration systems for default resource preferences
Error reporting for resource allocation issues
Includes¶
algorithm
cstdlib
functional
string