Slurm Workload Manager (SLURM)
Efficient Cluster Management for Composable Computing with FabreXTM and Slurm
Slurm (Simple Linux Utility for Resource Management) is the industry’s leading open-source, highly-scalable, and configurable cluster management and job scheduling system for Linux and Unix-like kernels. Slurm is the workload manager on about 60% of the TOP500 supercomputers around the world. It is designed to be highly efficient and fault-tolerant, making it well-suited for use in large-scale computing environments such as supercomputers and high-performance computing clusters. Slurm is widely used in scientific and academic computing, and is supported by a large and active community of developers and users.
- Highly scalable and configurable
- Optimized resource allocations for network topology and on-node topology
- Sophisticated multifactor job prioritization algorithms
- Real-time accounting down to the task level (identify specific tasks with high CPU or memory usage)
In a joint effort of many companies and organizations around the world, Slurm development is led by SchedMD.
GigaIO’s Fabrex Integration with Slurm
Thanks to GigaIO’s open ecosystem platform, FabreX integrates seamlessly with Slurm to add composable storage and GPUs as Slurm resources. These resources can then be dynamically allocated to servers matched to workflow demands.