Performance Optimization
Keep CastellanAI running smoothly with minimal impact on your systems.
Quick Performance Checks
If you're experiencing performance issues, start with these quick checks:
| Symptom | Quick Fix |
|---|---|
| Dashboard loading slowly | Narrow the time range filter (use "Last 1 Hour") |
| Agent using too much CPU | Enable Smart Filtering in agent settings |
| Events appearing delayed | Check agent connectivity status |
Dashboard Performance
Slow Page Loading
What you see: Dashboard pages take more than 5 seconds to load.
How to fix:
-
Use shorter time ranges
- Select "Last 1 Hour" or "Last 24 Hours" instead of longer ranges
- Large time ranges mean more data to process
-
Apply filters
- Filter by severity (e.g., "High and Critical only")
- Filter by event type or specific devices
- Fewer events = faster loading
-
Clear browser cache
- Old cached data can slow things down
- Press Ctrl+F5 to refresh and clear cache
-
Reduce auto-refresh frequency
- Go to Settings → Display
- Change auto-refresh from 5 seconds to 30 seconds or manual
Search Taking Too Long
What you see: Event searches take more than 10 seconds.
How to fix:
- Narrow your search criteria - Add more specific filters
- Use shorter date ranges - Search within last 24 hours first
- Search during off-peak hours - If possible, run large searches when system load is lower
Agent Performance
Agent Using Too Much CPU
What you see: Agent process consuming more than 15% CPU consistently.
How to fix:
- Go to Devices → Agents in the Portal
- Select the affected agent
- Open the Settings tab
- Make these changes:
| Setting | Recommended Value |
|---|---|
| Smart Filtering | Enabled |
| Monitoring Mode | Balanced (not Comprehensive) |
Agent Using Too Much Memory
What you see: Agent using more than 500 MB of RAM.
How to fix:
-
Update to the latest agent version
- Go to Devices → Updates
- Install any available agent updates
-
Restart the agent service
- This clears any accumulated memory usage
- The agent will restart automatically if configured
-
Reduce monitoring scope
- Disable monitoring for event channels you don't need
- Focus on Security events for most systems
Events Taking Too Long to Appear
What you see: Events take more than 30 seconds to show up in dashboard.
How to fix:
-
Check agent connectivity
- Go to Devices → Agents
- Verify the agent shows as "Online" (green)
-
Check network conditions
- High network latency or packet loss delays event transmission
- Contact your IT team to check network health
-
Verify smart filtering settings
- Low-priority events are batched and sent periodically
- Critical events should appear within seconds
Optimizing for Your Environment
Production Servers
For servers running critical applications:
| Recommendation | Why |
|---|---|
| Enable Smart Filtering | Reduces CPU usage significantly |
| Focus on Security channel | Other channels may be noisy |
| Monitor during maintenance windows | Review settings during low-usage periods |
Workstations
For end-user devices:
| Recommendation | Why |
|---|---|
| Use "Balanced" monitoring mode | Less impact on user experience |
| Monitor Security events only | Reduces data volume |
| Schedule updates for off-hours | Avoid interrupting users |
High-Volume Environments
For systems generating many events:
| Recommendation | Why |
|---|---|
| Increase batch interval | Reduces network overhead |
| Enable aggressive filtering | Drops low-value events earlier |
| Consider Enterprise tier | Higher rate limits and retention |
Monitoring System Health
Check overall system performance from the dashboard:
- Go to Settings → System Status
- Review key metrics:
| Metric | Healthy | Warning | Critical |
|---|---|---|---|
| Agent CPU Usage | Under 5% | 5-15% | Over 15% |
| Event Processing Delay | Under 10 sec | 10-30 sec | Over 30 sec |
| Agent Memory | Under 200 MB | 200-500 MB | Over 500 MB |
What's Next?
- Agent Health Monitoring - Monitor agent status and performance
- Agent Settings - Configure agent behavior
- Connectivity Issues - Resolve connection problems