“Sinfo” is a command-line tool that is part of the Slurm workload manager, which is a highly scalable and flexible job scheduling and management system used in high-performance computing (HPC) environments. The purpose of “sinfo” is to provide users with information about the nodes and partitions that are managed by the Slurm scheduler.
“Sinfo” can be used to view a wide range of information about the nodes and partitions, including the number of nodes available, their state, and the amount of memory and CPU resources available on each node. This information can be useful for users who need to schedule and manage their computational workloads efficiently and effectively.
Some of the options available with “sinfo” include the ability to display information about specific partitions or nodes, to filter the output based on various criteria, and to sort the output based on different fields. This allows users to quickly and easily identify the resources that are available to them and to plan their workloads accordingly.
sinfo Command Examples
1. Show a quick summary overview of the cluster:
# sinfo --summarize
2. View the detailed status of all partitions across the entire cluster:
# sinfo
3. View the detailed status of a specific partition:
# sinfo --partition partition_name
4. View information about idle nodes:
# sinfo --states idle
5. Summarise dead nodes:
# sinfo --dead
6. List dead nodes and the reasons why:
# sinfo --list-reasons
Summary
In addition to “sinfo”, the Slurm workload manager also includes several other command-line tools that are commonly used in HPC environments. These include “squeue”, which allows users to view information about jobs that are currently running or waiting in the queue, and “sbatch”, which is used to submit batch jobs to the scheduler.
Overall, “sinfo” is a powerful and useful tool for users who need to manage and schedule computational workloads in HPC environments. By providing detailed information about the nodes and partitions that are available, users can optimize their use of system resources and ensure that their workloads are executed efficiently and effectively.