Features - NodeGuarder

eBPF Cron Monitoring

Stop guessing if your backups actually ran. NodeGuarder uses eBPF to hook directly into the kernel, detecting cron job executions, exit codes, and timeouts without any wrapper scripts.

Zero Touch: No need to modify crontab or use `curl` pings.
Precise Detection: Captures exact exit codes via `sched_process_exit`.
Timeout Alerts: Know instantly if a job hangs (Long Running).
Auto-Discovery: Automatically finds and tracks new cron jobs.

                        Why eBPF?

                        Legacy tools rely on parsing log files, which is brittle and slow. eBPF allows us to observe the
                        actual process execution in real-time with zero overhead.
                    

// Dashboard View

backup.sh PID: 1042

● Failed (Exit Code 1)
Ran: 04:00:01 - 04:00:03 (2s)

cleanup.py PID: 2155

● Success (Exit Code 0)
Ran: 04:05:00 - 04:05:45 (45s)

sync_db.sh PID: 3402

● Timeout (> 300s limit)
Started: 04:10:00 (Running for 12m 30s...)

// Metrics collected

CPU Usage: 42% (Normalizing)
Memory: 4.2GB / 16GB
Disk: 45% used
Load Avg: 0.95, 1.10, 0.85
Uptime: 45 days

Status: Recovering (Window: 20s remaining)
↓
Status: Healthy

Health & Performance

Keep a finger on the pulse of your infrastructure. We track valid metrics and alert you when sustained pressure threatens stability.

Smart Thresholds: Define Warning & Critical levels for CPU, RAM, and Disk.
False Positive Protection: "Sustain Duration" ensures transient spikes don't wake you up at 3 AM.
Smart Recovery: Configurable "Stability Window" prevents flapping alerts by ensuring a server is truly healthy before sending a "Resolved" notification.
Offline Watchdog: Instant alerts if a server stops hearbeating for more than 2 minutes (default, configurable).

Configuration Drift (FIM)

Unplanned changes are the root cause of most outages. NodeGuarder monitors your `/etc` directory (or any path) for unauthorized modifications.

File Integrity: Checksum-based tracking of config files.
Granular Control: Ignore temporary files (`*.tmp`, `*.bak`) to reduce noise.
Audit Trail: Know exactly when a file changed.

// Alert: Drift Detected

/etc/nginx/nginx.conf
● MODIFIED
Time: 14:02:23 (Checksum mismatch)

/etc/cron.d/backdoor
● CREATED
Time: 14:05:01 (New file detected)

/etc/hosts.allow
● DELETED
Time: 14:10:15 (File removed)

// #alerts-channel

NodeGuarder Bot 10:42 AM
[CRITICAL] Server 'prod-db-01' High CPU Load (98%)

NodeGuarder BOT
[WARN] Cron job 'backup.sh' timed out (> 300s)

NodeGuarder
[RESOLVED] Server 'web-02' is back online.

Automatic notification of alerts

Don't stare at dashboards all day. NodeGuarder integrates with the tools you already use, sending instant notifications when something needs your attention.

Multi-Channel Support: Seamless integration with Slack, Microsoft Teams, and Discord.
Email Alerts: Robust SMTP support (including StartTLS) for traditional email notifications.
Smart Routing: Configure different alert severities (Critical, Warning) to trigger different notification channels.
No More Flapping: "Resolved" notifications are only sent after the server has initialized and remained stable for a configurable duration.
Rich Context: Alerts include the server name, error details, and relevant metrics so you can act immediately.

Ready to take control?

Join the community of sysadmins who sleep properly at night.

Get Started Now