Command Reference

Complete reference of all S9S commands, shortcuts, and operations for efficient cluster management.

📋 Command Categories

Navigation Commands

Command	Description	Shortcut
`:view jobs`	Switch to jobs view	`1`
`:view nodes`	Switch to nodes view	`2`
`:view users`	Switch to users view	`3`
`:view partitions`	Switch to partitions view	`4`
`:view dashboard`	Switch to dashboard	`0`
`:help`	Show help	`?`
`:quit`	Exit S9S	`q`

Job Management Commands

Command	Description	Example
`:submit`	Submit a new job	`:submit script.sh --cores=4`
`:cancel`	Cancel job(s)	`:cancel 12345`
`:hold`	Hold job(s)	`:hold 12345 --reason="Debug"`
`:release`	Release held job(s)	`:release 12345`
`:requeue`	Requeue job(s)	`:requeue 12345`
`:priority`	Change job priority	`:priority 12345 --priority=1000`

Node Management Commands

Command	Description	Example
`:drain`	Drain node(s)	`:drain node001 --reason="Maintenance"`
`:resume`	Resume drained node(s)	`:resume node001`
`:ssh`	SSH to node	`:ssh node001`
`:reboot`	Reboot node	`:reboot node001`

Filtering Commands

Command	Description	Example
`/`	Quick filter	`/user:alice`
`:filter`	Advanced filter	`:filter "state:RUNNING nodes:>4"`
`:clear`	Clear all filters	`:clear`
`:save-filter`	Save current filter	`:save-filter my-jobs`

Export Commands

Command	Description	Example
`:export`	Export data	`:export csv --output=jobs.csv`
`:report`	Generate report	`:report utilization --period=week`

🎯 Detailed Command Reference

Job Commands

:submit - Submit Jobs


:submit <script> [options]

Options:
  --cores=N           Number of CPU cores
  --mem=SIZE          Memory requirement (e.g., 32GB)
  --time=DURATION     Time limit (e.g., 2:00:00)
  --partition=NAME    Partition name
  --gpus=N           Number of GPUs
  --array=RANGE      Job array (e.g., 1-100)
  --dependency=TYPE:ID Job dependency
  --name=NAME        Job name
  --account=ACCOUNT  Account for billing

Examples:
  :submit analysis.sh --cores=16 --mem=64GB --time=4:00:00
  :submit ml_train.py --gpus=2 --partition=gpu
  :submit job_array.sh --array=1-100 --cores=4

:cancel - Cancel Jobs


:cancel <job_ids> [options]

Options:
  --force            Force cancellation
  --reason=TEXT      Cancellation reason
  --signal=SIGNAL    Signal to send (default: SIGTERM)

Examples:
  :cancel 12345
  :cancel 12345,12346,12347
  :cancel job[001-100]
  :cancel --force 12345 --reason="Emergency maintenance"

:hold - Hold Jobs


:hold <job_ids> [options]

Options:
  --reason=TEXT      Hold reason
  --user-hold        User hold (can release)
  --system-hold      System hold (admin required)

Examples:
  :hold 12345 --reason="Debugging required"
  :hold /state:PENDING --user-hold

:release - Release Jobs


:release <job_ids>

Examples:
  :release 12345
  :release /state:HELD user:alice

:requeue - Requeue Jobs


:requeue <job_ids> [options]

Options:
  --hold             Hold after requeue
  --reason=TEXT      Requeue reason

Examples:
  :requeue 12345
  :requeue /state:FAILED --hold

:priority - Change Priority


:priority <job_ids> --priority=VALUE

Examples:
  :priority 12345 --priority=1000
  :priority /user:alice --priority=+100
  :priority /state:PENDING --priority=500

Node Commands

:drain - Drain Nodes


:drain <nodes> [options]

Options:
  --reason=TEXT      Drain reason
  --timeout=DURATION Max time to wait for jobs
  --force            Force drain (kill jobs)

Examples:
  :drain node001 --reason="Hardware upgrade"
  :drain node[001-010] --timeout=2h
  :drain /features:gpu --force --reason="Emergency"

:resume - Resume Nodes


:resume <nodes>

Examples:
  :resume node001
  :resume node[001-010]
  :resume /state:DRAIN

:ssh - SSH to Nodes


:ssh <node> [command] [options]

Options:
  --user=USERNAME    SSH username
  --key=KEYFILE      SSH key file
  --port=PORT        SSH port

Examples:
  :ssh node001
  :ssh node001 "uptime"
  :ssh node001 --user=admin "sudo reboot"

:reboot - Reboot Nodes


:reboot <nodes> [options]

Options:
  --force            Force reboot
  --delay=SECONDS    Delay before reboot
  --reason=TEXT      Reboot reason

Examples:
  :reboot node001
  :reboot node[001-010] --delay=300
  :reboot /state:MAINT --force

Filtering Commands

/ - Quick Filter


/<filter_expression>

Examples:
  /user:alice                    # Jobs by user alice
  /state:RUNNING                 # Running jobs
  /partition:gpu state:PENDING   # Pending GPU jobs
  /nodes:>4 time:>2h            # Large, long-running jobs

:filter - Advanced Filter


:filter "<complex_filter>"

Examples:
  :filter "user:alice,bob state:RUNNING partition:gpu"
  :filter "submitted:>1h runtime:<30m"
  :filter "priority:>1000 OR qos:high"

:save-filter - Save Filters


:save-filter <name> [filter]

Examples:
  :save-filter my-running-jobs "/user:alice state:RUNNING"
  :save-filter gpu-jobs "/partition:gpu"
  :save-filter # Save current filter

:load-filter - Load Filters


:load-filter <name>

Examples:
  :load-filter my-running-jobs
  :load-filter gpu-jobs

Export Commands

:export - Export Data


:export <format> [options]

Formats:
  csv, json, excel, html, xml

Options:
  --output=FILE      Output filename
  --fields=LIST      Fields to export
  --filter=EXPR      Filter expression
  --selected         Export selected items only

Examples:
  :export csv --output=jobs.csv
  :export json --fields=JobID,User,State
  :export excel --selected
  :export csv --filter="user:alice" --output=alice_jobs.csv

:report - Generate Reports


:report <type> [options]

Types:
  utilization, efficiency, summary, costs, compliance

Options:
  --period=DURATION  Time period (day, week, month, year)
  --format=FORMAT    Output format (pdf, html, csv)
  --email=ADDRESS    Email report
  --users=LIST       Specific users
  --partitions=LIST  Specific partitions

Examples:
  :report utilization --period=month --format=pdf
  :report efficiency --users=alice,bob
  :report costs --partitions=gpu [email protected]

View Commands

:view - Switch Views


:view <view_name>

Views:
  jobs, nodes, users, partitions, dashboard, performance

Examples:
  :view jobs
  :view nodes
  :view dashboard

:columns - Configure Columns


:columns <action> [column_name]

Actions:
  show, hide, add, remove, reset

Examples:
  :columns show Priority
  :columns hide Account  
  :columns add GPUs
  :columns reset

:sort - Sort Data


:sort <column> [direction]

Direction:
  asc, desc

Examples:
  :sort JobID desc
  :sort SubmitTime asc
  :sort Priority desc

Configuration Commands

:config - Configuration Management


:config <action> [key] [value]

Actions:
  get, set, list, reset, validate

Examples:
  :config get refresh_interval
  :config set theme dark
  :config list
  :config validate

:theme - Theme Management


:theme <theme_name>

Themes:
  dark, light, terminal, high-contrast

Examples:
  :theme dark
  :theme terminal

System Commands

:status - System Status


:status [component]

Components:
  cluster, scheduler, storage, network

Examples:
  :status
  :status cluster
  :status scheduler

:log - View Logs


:log [options]

Options:
  --lines=N          Number of lines
  --follow           Follow log output
  --level=LEVEL      Log level filter
  --component=COMP   Component filter

Examples:
  :log --lines=100
  :log --follow --level=error
  :log --component=scheduler

:refresh - Refresh Data


:refresh [view]

Examples:
  :refresh           # Refresh current view
  :refresh jobs      # Refresh jobs data
  :refresh nodes     # Refresh nodes data

⌨️ Keyboard Shortcuts

Global Shortcuts

Key	Action
`q`	Quit
`?`	Help
`r`	Refresh
`Ctrl+C`	Interrupt/Cancel
`Esc`	Clear/Cancel

Navigation Shortcuts

Key	Action
`0`	Dashboard view
`1`	Jobs view
`2`	Nodes view
`3`	Users view
`4`	Partitions view
`Tab`	Next view
`Shift+Tab`	Previous view

View-Specific Shortcuts

Jobs View:

Key	Action
`c`	Cancel job
`h`	Hold job
`r`	Release job
`k`	Kill job
`d`	Show details
`s`	SSH to job nodes
`o`	Show output
`l`	Show logs

Nodes View:

Key	Action
`d`	Drain node
`r`	Resume node
`s`	SSH to node
`m`	Maintenance mode
`i`	Node info
`j`	Jobs on node

Selection Shortcuts:

Key	Action
`Space`	Toggle selection
`v`	Visual selection mode
`V`	Visual line mode
`Ctrl+A`	Select all
`Ctrl+D`	Deselect all

Filtering Shortcuts

Key	Action
`/`	Quick filter
`n`	Next match
`N`	Previous match
`Ctrl+/`	Clear filter

Export Shortcuts

Key	Action
`Ctrl+E`	Export CSV
`Ctrl+Shift+E`	Export JSON
`Alt+E`	Export Excel
`Ctrl+P`	Print/PDF

🔧 Advanced Usage

Command Chaining


# Chain multiple commands
:filter "state:FAILED" ; :cancel --selected ; :clear

# Conditional execution
:submit job.sh --cores=16 && :priority $JOB_ID --priority=1000

# Pipe operations
:export json --fields=JobID,User | :import --target=database

Variables and Substitution


# Use environment variables
:submit $HOME/scripts/analysis.sh --account=$SLURM_ACCOUNT

# Job ID substitution
:cancel $LAST_JOB_ID
:ssh node001 "kill $JOB_PID"

# Date/time substitution
:export csv --output=jobs_$(date +%Y%m%d).csv

Scripting Commands


# Script execution
:script /path/to/commands.s9s

# Batch mode
:batch < batch_commands.txt

# Interactive scripting
:begin-script
:filter "user:alice state:RUNNING"
:priority --selected --priority=500
:end-script

🚨 Emergency Commands

Emergency Stop


:emergency-stop [options]

Options:
  --reason=TEXT      Emergency reason
  --notify           Send notifications
  --graceful         Graceful shutdown

Examples:
  :emergency-stop --reason="Power outage" --notify

Emergency Evacuation


:evacuate <nodes> [options]

Options:
  --timeout=DURATION Evacuation timeout
  --force            Force evacuation
  --destination=NODE Target nodes

Examples:
  :evacuate rack01 --timeout=30m
  :evacuate /features:gpu --force

💡 Command Tips

Efficiency Tips

Use tab completion - Most commands support tab completion
Learn shortcuts - Master keyboard shortcuts for speed
Save filters - Save commonly used filters
Use aliases - Create command aliases for frequent operations
Batch operations - Use visual selection for bulk actions

Best Practices

Be specific - Use precise filters and job IDs
Confirm destructive actions - Always verify before canceling jobs
Use dry-run mode - Test commands with
```
--dry-run
```
when available
Document reasons - Always provide reasons for administrative actions
Check permissions - Ensure you have necessary permissions

🔍 Command Help

Getting Help


:help                    # General help
:help <command>         # Command-specific help
:help shortcuts         # Keyboard shortcuts
:help examples          # Command examples
:man <command>          # Detailed manual page

Command Discovery


:commands               # List all commands
:commands --category=job # Commands by category
:search <keyword>       # Search commands
:recent                 # Recently used commands

🚀 Next Steps

Practice with Job Management commands
Learn Advanced Filtering syntax
Explore Batch Operations for efficiency
Master SSH Integration commands