Skip to main content
The /logs command displays OpenClaw service logs from a deployed instance. It fetches logs from the systemd journal and provides real-time monitoring capabilities.

​
Usage

/logs

​
Log Viewing Process

​
1. Select Deployment

Choose which deployment to view logs from: 
/logs - Select a deployment

┌────────────────────────────────────────────┐
│ > production-claude    203.0.113.45        │
│   staging-test        198.51.100.23        │
└────────────────────────────────────────────┘

Arrow keys to select | Enter to view logs | Esc to go back

​
2. Loading Logs

Loading logs...
Fetching OpenClaw logs from server...

​
3. Log Display

View the most recent 200 log lines: 
Logs: production-claude | Auto: OFF | Fetched: 2:45:30 PM

┌────────────────────────────────────────────────────────────────┐
│ Feb 28 14:30:15 OpenClaw agent starting...                     │
│ Feb 28 14:30:16 Connected to Anthropic API                     │
│ Feb 28 14:30:17 Telegram bot initialized                       │
│ Feb 28 14:30:18 Gateway listening on port 18789                │
│ Feb 28 14:35:22 Received message from user 123456789           │
│ Feb 28 14:35:23 Processing command: /help                      │
│ Feb 28 14:35:25 Response sent                                  │
│ Feb 28 14:40:01 Health check: OK                               │
│ Feb 28 14:45:15 Memory usage: 456MB / 2GB                      │
└────────────────────────────────────────────────────────────────┘

R: Refresh | A: Toggle auto-refresh | Esc: Back | Showing last 200 lines

​
Log Features

​
Color-Coded Levels

Logs are automatically color-coded by severity:
  • Error (Red): Critical errors and failures
  • Warning (Yellow): Warnings and potential issues
  • Info (Default): Normal operational messages
  • Debug (Gray): Detailed debugging information

​
Manual Refresh

Press R to fetch fresh logs: 
Logs: production-claude | Auto: OFF | Fetched: 2:46:15 PM

[Latest logs appear]

​
Auto-Refresh Mode

Press A to enable automatic refresh every 5 seconds: 
Logs: production-claude | Auto: ON (5s) | Fetched: 2:46:20 PM

[Logs automatically update every 5 seconds]
Press A again to disable: 
Logs: production-claude | Auto: OFF | Fetched: 2:46:45 PM

​
Log Format

Logs follow the systemd journal format: 
Feb 28 14:30:15 hostname openclaw[pid]: message
ClawControl parses and displays: 
Feb 28 14:30:15 message

​
Example Log Output

​
Startup Logs

Feb 28 10:45:00 OpenClaw agent starting (v1.2.3)
Feb 28 10:45:01 Environment: production
Feb 28 10:45:01 Configuration loaded from /root/.openclaw/config.json
Feb 28 10:45:02 AI Provider: anthropic (claude-sonnet-4-20250514)
Feb 28 10:45:03 Channel: telegram (bot @my_openclaw_bot)
Feb 28 10:45:04 Gateway server started on port 18789
Feb 28 10:45:05 OpenClaw is ready

​
Normal Operation

Feb 28 11:23:45 [Telegram] Message received from user 123456789
Feb 28 11:23:46 [Agent] Processing user request
Feb 28 11:23:47 [AI] Sending prompt to Anthropic API
Feb 28 11:23:49 [AI] Received response (156 tokens)
Feb 28 11:23:50 [Agent] Executing tool: bash
Feb 28 11:23:51 [Tool] Command completed successfully
Feb 28 11:23:52 [Telegram] Response sent to user

​
Error Logs

Feb 28 12:15:30 ERROR: Failed to connect to AI provider
Feb 28 12:15:30 Error: ECONNREFUSED
Feb 28 12:15:30 Retrying in 5 seconds...
Feb 28 12:15:35 Connection established

​
Warning Logs

Feb 28 13:45:12 WARN: High memory usage detected (1.8GB / 2GB)
Feb 28 13:45:12 WARN: Consider upgrading server size
Feb 28 13:45:15 Memory usage normal (1.2GB / 2GB)

​
Terminal Session Example

> /logs

/logs - Select a deployment

┌────────────────────────────────────────────┐
│ > production-claude    203.0.113.45        │
└────────────────────────────────────────────┘

[Enter]

Loading logs...

Logs: production-claude | Auto: OFF | Fetched: 2:47:10 PM

┌────────────────────────────────────────────────────────────────┐
│ Feb 28 14:30:15 OpenClaw agent starting...                     │
│ Feb 28 14:30:16 Connected to Anthropic API                     │
│ Feb 28 14:35:22 Received message from user 123456789           │
│ Feb 28 14:40:01 Health check: OK                               │
└────────────────────────────────────────────────────────────────┘

[Press R]

Fetching logs...

Logs: production-claude | Auto: OFF | Fetched: 2:47:15 PM

[Updated logs appear]

[Press A]

Logs: production-claude | Auto: ON (5s) | Fetched: 2:47:20 PM

[Logs refresh automatically every 5 seconds]

[Press Esc]

> _

​
Keyboard Controls

  • R - Refresh logs manually
  • A - Toggle auto-refresh (5 second interval)
  • Esc - Return to home
  • ↑/↓ - Scroll through logs (in selection view)
  • Enter - Load logs for selected deployment

​
Empty State

If no deployed instances exist: 
/logs - View deployment logs

┌────────────────────────────────────────────┐
│ No deployed instances found!               │
│ Deploy an instance first with /deploy      │
└────────────────────────────────────────────┘

Press any key to return to home

​
Error Handling

​
Connection Failed

Error Loading Logs

┌────────────────────────────────────────────┐
│ Failed to fetch logs: Connection timeout   │
└────────────────────────────────────────────┘

Press any key to go back
Solutions:
  1. Check server is running: /status
  2. Verify SSH connectivity: /ssh
  3. Check firewall rules

​
Service Not Running

Feb 28 15:00:00 ERROR: Service openclaw not found
Solutions:
  1. SSH into server: /ssh
  2. Check service status: systemctl status openclaw
  3. Start service: systemctl start openclaw

​
Advanced Log Analysis

For detailed log analysis, use SSH:

​
View Full Logs

root@server:~# journalctl -u openclaw --no-pager

​
Filter by Time

root@server:~# journalctl -u openclaw --since "1 hour ago"
root@server:~# journalctl -u openclaw --since "2026-02-28 14:00:00"

​
Filter by Level

root@server:~# journalctl -u openclaw -p err
# Only errors

root@server:~# journalctl -u openclaw -p warning
# Warnings and above

​
Follow Logs in Real-Time

root@server:~# journalctl -u openclaw -f
# Similar to 'tail -f'

​
Export Logs

root@server:~# journalctl -u openclaw > openclaw.log
root@server:~# journalctl -u openclaw --since "today" > openclaw-today.log

​
Search Logs

root@server:~# journalctl -u openclaw | grep ERROR
root@server:~# journalctl -u openclaw | grep "user 123456789"

​
Log Rotation

Systemd automatically rotates logs. Configure retention: 
root@server:~# vim /etc/systemd/journald.conf

SystemMaxUse=500M
SystemMaxFileSize=50M
MaxRetentionSec=7day

root@server:~# systemctl restart systemd-journald

​
Tips

Use auto-refresh mode when actively troubleshooting issues. Turn it off when done to reduce server load.
For long-term log analysis, use /ssh and journalctl’s advanced filtering options rather than viewing logs in ClawControl.
ClawControl displays the last 200 lines by default. For complete logs, use SSH and journalctl directly.

​
Common Log Patterns

​
Successful User Interaction

[Telegram] Message received
[Agent] Processing
[AI] Sending prompt
[AI] Received response
[Tool] Executing
[Tool] Completed
[Telegram] Response sent

​
API Rate Limit

ERROR: Rate limit exceeded
WARN: Waiting 60 seconds before retry
Retrying request...
Connection established

​
Memory Issues

WARN: High memory usage (1.8GB / 2GB)
WARN: Running garbage collection
Memory freed: 400MB
Memory usage normal

​
Next Steps

SSH Command

Advanced log analysis

Status Command

Check deployment health

Dashboard Command

Web-based monitoring

Deploy Command

Restart failed services

Build docs developers (and LLMs) love

Get started for freeTalk to us