Skip to main content
ClawControl provides built-in log viewing to help you monitor and troubleshoot your OpenClaw deployments without leaving the CLI.

​
Quick Log Access

1

Open logs viewer

From ClawControl’s main menu:
/logs
Select a deployed instance:
/logs - Select a deployment

> my-openclaw-server     159.89.123.45
  prod-server            142.93.45.67

Arrow keys to select | Enter to view logs | Esc to go back
2

View logs

Press Enter to load logs. ClawControl connects via SSH and fetches the last 200 log lines:
Logs: my-openclaw-server | Auto: OFF | Fetched: 3:45:23 PM

Jan 15 14:45:23 OpenClaw agent started
Jan 15 14:45:24 Connected to Telegram
Jan 15 14:45:25 AI provider: openrouter
Jan 15 14:45:26 Model: moonshotai/kimi-k2.5
Jan 15 14:45:27 Gateway listening on port 18789
Jan 15 14:46:01 New message from user 123456789
Jan 15 14:46:03 Generated response (245 tokens)
Jan 15 14:46:04 Sent response to Telegram

R: Refresh | A: Toggle auto-refresh | Esc: Back
Showing last 200 lines
3

Refresh logs

Press R to manually refresh and fetch new log entries.Press A to enable auto-refresh (every 5 seconds):
Logs: my-openclaw-server | Auto: ON (5s) | Fetched: 3:45:28 PM

​
Understanding Log Entries

OpenClaw logs use journalctl format: 
Jan 15 14:45:23 hostname openclaw[1234]: OpenClaw agent started
│              │        │        │      │
│              │        │        │      └─ Log message
│              │        │        └─ Process ID
│              │        └─ Service name
│              └─ Hostname
└─ Timestamp (month day time)

​
Log Levels

Logs are color-coded by severity:
  • Info (white): Normal operations 
    Jan 15 14:45:23 Gateway listening on port 18789
  • Debug (gray): Detailed diagnostic information 
    Jan 15 14:45:24 DEBUG: Initializing Telegram client
  • Warning (yellow): Potential issues 
    Jan 15 14:46:15 Warning: Rate limit approaching
  • Error (red): Failures and errors 
    Jan 15 14:47:32 Error: Failed to connect to AI provider

​
Common Log Patterns

​
Startup Sequence

Normal OpenClaw startup: 
Jan 15 14:45:23 OpenClaw agent started
Jan 15 14:45:24 Loading configuration from /root/.openclaw/config.json
Jan 15 14:45:24 AI Provider: openrouter
Jan 15 14:45:24 Model: moonshotai/kimi-k2.5
Jan 15 14:45:25 Connecting to Telegram...
Jan 15 14:45:26 Connected to Telegram as @my_bot
Jan 15 14:45:27 Gateway server listening on 0.0.0.0:18789
Jan 15 14:45:27 OpenClaw ready

​
Message Processing

Jan 15 15:30:12 New message from user 123456789: "Hello"
Jan 15 15:30:13 Generating response...
Jan 15 15:30:14 OpenRouter request: moonshotai/kimi-k2.5
Jan 15 15:30:16 Response received (142 tokens)
Jan 15 15:30:16 Sent response to Telegram

​
Error Examples

API Key Invalid: 
Jan 15 15:45:23 Error: AI provider authentication failed
Jan 15 15:45:23 Check your API key configuration
Network Error: 
Jan 15 16:12:45 Error: Failed to connect to api.openrouter.ai
Jan 15 16:12:45 Retrying in 5 seconds...
Telegram Error: 
Jan 15 17:23:11 Error: Telegram webhook failed
Jan 15 17:23:11 Bot token may be invalid

​
Log Filtering

View specific log types directly via SSH:

​
Errors Only

ssh root@YOUR_SERVER_IP "journalctl -u openclaw | grep -i error"

​
Warnings and Errors

ssh root@YOUR_SERVER_IP "journalctl -u openclaw | grep -iE 'warn|error'"

​
Last Hour

ssh root@YOUR_SERVER_IP "journalctl -u openclaw --since '1 hour ago'"

​
Date Range

ssh root@YOUR_SERVER_IP "journalctl -u openclaw --since '2026-01-15 14:00' --until '2026-01-15 16:00'"

​
Troubleshooting with Logs

​
Bot Not Responding

1

Check if service is running

Look for startup messages:
Jan 15 14:45:27 OpenClaw ready
If missing, check for errors or restart messages.
2

Verify Telegram connection

Look for:
Jan 15 14:45:26 Connected to Telegram as @my_bot
If missing, check bot token:
Error: Telegram authentication failed
3

Check for access control errors

If users are blocked:
Jan 15 15:30:12 Rejected message from unauthorized user: 987654321
Update telegramAllowFrom in configuration.

​
AI Responses Failing

1

Check AI provider connection

Jan 15 15:30:14 OpenRouter request: moonshotai/kimi-k2.5
Jan 15 15:30:16 Response received (142 tokens)
If you see errors:
Error: AI provider authentication failed
2

Verify API key

From SSH:
cat ~/.openclaw/.env
Check the AI provider API key is correct.
3

Check rate limits

Warning: Rate limit approaching
Error: Rate limit exceeded, retry after 60s
Wait or upgrade your AI provider plan.

​
High Memory Usage

1

Check for memory warnings

Warning: Memory usage at 85%
Error: JavaScript heap out of memory
2

Verify server size

From /status, check your deployment:
  • Minimum: 2GB RAM
  • Recommended: 4GB RAM for production
3

Restart service

Via SSH:
systemctl restart openclaw

​
Advanced Log Analysis

​
Export Logs to File

Via SSH: 
journalctl -u openclaw > openclaw-logs.txt
Download to local machine: 
scp root@YOUR_SERVER_IP:openclaw-logs.txt .

​
Search Logs

Search for specific terms: 
journalctl -u openclaw | grep "search term"
Example - find all errors: 
journalctl -u openclaw | grep -i error

​
Count Log Entries

Count errors in last hour: 
journalctl -u openclaw --since "1 hour ago" | grep -i error | wc -l

​
Monitor in Real-Time

Follow logs as they happen: 
journalctl -u openclaw -f
Press Ctrl+C to stop.

​
Log Rotation

OpenClaw uses systemd’s journald, which automatically rotates logs:
  • Default retention: 1 month
  • Max size: 4GB
  • Location: /var/log/journal/

​
View Log Disk Usage

journalctl --disk-usage
Output: 
Archived and active journals take up 245.6M in the file system.

​
Clean Old Logs

Remove logs older than 7 days: 
journalctl --vacuum-time=7d
Limit total size to 500MB: 
journalctl --vacuum-size=500M

​
Performance Monitoring

​
Track Response Times

Search for response time logs: 
journalctl -u openclaw | grep "Response received"
Example output: 
Jan 15 15:30:16 Response received (142 tokens, 2.3s)
Jan 15 15:35:22 Response received (389 tokens, 4.1s)
Jan 15 15:42:18 Response received (95 tokens, 1.2s)

​
Monitor Message Volume

Count messages per hour: 
journalctl -u openclaw --since "1 hour ago" | grep "New message" | wc -l

​
Track Errors Over Time

Errors in last 24 hours by hour: 
for i in {0..23}; do
  echo -n "Hour -$i: "
  journalctl -u openclaw --since "$i hours ago" --until "$((i-1)) hours ago" \
    | grep -i error | wc -l
done

​
Alerting and Monitoring

For production deployments, consider setting up external monitoring:
  • Uptime monitoring: UptimeRobot, Pingdom
  • Log aggregation: Papertrail, Loggly
  • Application monitoring: New Relic, Datadog
  • Error tracking: Sentry, Rollbar

​
Simple Health Check Script

Create a monitoring script on your server: 
#!/bin/bash
# health-check.sh

# Check if OpenClaw is running
if ! systemctl is-active --quiet openclaw; then
  echo "ALERT: OpenClaw is not running"
  # Send notification (email, Slack, etc.)
  exit 1
fi

# Check for recent errors
ERRORS=$(journalctl -u openclaw --since "5 minutes ago" | grep -i error | wc -l)
if [ "$ERRORS" -gt 5 ]; then
  echo "ALERT: $ERRORS errors in last 5 minutes"
  exit 1
fi

echo "OK: OpenClaw is healthy"
exit 0
Run via cron every 5 minutes: 
crontab -e
Add: 
*/5 * * * * /root/health-check.sh

​
Best Practices

Regular log review:
  • Check logs daily for new deployments
  • Weekly review for production systems
  • Investigate any unexpected errors immediately
What to watch for:
  1. Startup errors: Service fails to start
  2. Authentication failures: API keys invalid or expired
  3. Rate limiting: Approaching or exceeding limits
  4. Network issues: Connection timeouts
  5. Memory warnings: Approaching system limits
When to take action:
  • Immediate: Service down, authentication failures
  • Within 1 hour: Rate limiting, network errors
  • Within 24 hours: Memory warnings, performance degradation

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us