sdiag

Display Slurm controller diagnostic information

TLDR

Show scheduling diagnostic information (default mode)

$ sdiag

Show diagnostics sorted by RPC total run time

$ sdiag -t

Show diagnostics sorted by RPC average run time

$ sdiag -T

Reset performance counters (requires operator/admin privileges)

$ sdiag -r

Output diagnostics as JSON

$ sdiag --json

sdiag displays diagnostic information about slurmctld, the Slurm controller daemon. It shows performance metrics, scheduling statistics, RPC counters, and resource usage data.This is useful for monitoring cluster health, troubleshooting scheduling performance, and identifying bottlenecks in the Slurm controller.

PARAMETERS

-a, --all

Get and report information. This is the default mode of operation.

-h, --help

Print description of options and exit.

-i, --sort-by-id

Sort RPC data by message type ID and user ID.

-r, --reset

Reset scheduler and RPC counters to 0. Only supported for Slurm operators and administrators.

-t, --sort-by-time

Sort RPC data by total run time.

-T, --sort-by-time2

Sort RPC data by average run time.

--json

Output information as JSON.

--yaml

Output information as YAML.

-V, --version

Print version number and exit.

--usage

Print list of options and exit.

INSTALL

sudo apt install slurm-client

CAVEATS

Requires appropriate permissions to access Slurm controller data. The reset option requires operator or administrator privileges and affects all users' view of counters.

HISTORY

Part of Slurm workload manager, providing diagnostic tools for cluster administrators.