Sshare

sshare is a command-line tool in the SLURM workload manager, used for displaying historical resource usage and fair-share information for users and groups. This information is crucial for understanding how SLURM allocates resources based on past usage, aiding in efficient job scheduling.

Key Features

• Account/User Share Information: Displays the allocation of resources to users and accounts.
• Raw and Normalized Shares: Shows both actual and proportionate shares.
• Usage Data: Details resource consumption, aiding in monitoring and planning.
• Fair-share Factor: Indicates the priority based on usage and allocation. (Higher is better)

Usage

The basic syntax for sshare is:

<source lang="bash"> sshare [options] </source>

Common options include:

• -a or --all: Show all users/groups.
• -u or --user=[users]: Display information for specific user(s).
• -A or --account=[accounts]: Display information for specific account(s).
• -l or --long: Show detailed information.

Viewing Your Fair-Share Value

To view your fair-share value, simply run:

<source lang="bash"> sshare -u <username> </source>

Replace <username> with your actual username.

Checking Group Usage

To check the usage for a specific group:

<source lang="bash"> sshare -A <group_name> </source>

Replace <group_name> with the name of the group you want to check.

Detailed Information

For more detailed information, including fields like RawUsage, EffectvUsage, NormUsage, and FairShare:

<source lang="bash"> sshare -l </source>

Examples

1. View All User Usage:

<source lang="bash"> sshare -a </source> This command lists fair-share and usage information for all users.

1. Specific User in Detail:

<source lang="bash"> sshare -l -u jdoe </source> Replace jdoe with the username. This provides detailed usage information for user jdoe.

1. Multiple Users:

<source lang="bash"> sshare -u "user1,user2" </source> This shows the information for user1 and user2.

Conclusion

Understanding sshare outputs allows users to better comprehend how their and their group's historical usage impacts current and future resource allocation, leading to more efficient job submissions and scheduling on SLURM clusters.

See Also

• seff
• Slurm
• SLURM Job Management