multiverse/xcom-ultra/xcu-sentinel/src/lib.rs

#![deny(warnings)]
#![allow(dead_code)]
//! [TSM.ID].[11031972] -- Platform X Ecosystem
//! xcu-sentinel -- System Watchdog with Resource Monitoring
//! CPU/RAM/disk monitoring, threshold alerts, SLA enforcement

use std::collections::VecDeque;
use std::sync::{Arc, Mutex};
use std::time::SystemTime;

#[derive(Debug)]
pub enum SentinelError {
    ThresholdExceeded(String),
    MonitorFailed(String),
    ConfigError(String),
}

impl std::fmt::Display for SentinelError {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        match self {
            Self::ThresholdExceeded(e) => write!(f, "Threshold exceeded: {e}"),
            Self::MonitorFailed(e) => write!(f, "Monitor failed: {e}"),
            Self::ConfigError(e) => write!(f, "Config error: {e}"),
        }
    }
}

impl std::error::Error for SentinelError {}

#[derive(Debug, Clone)]
pub struct ResourceSnapshot {
    pub cpu_percent: f64,
    pub memory_used_mb: u64,
    pub memory_total_mb: u64,
    pub disk_used_percent: f64,
    pub open_connections: u32,
    pub timestamp: u64,
}

#[derive(Debug, Clone)]
pub struct AlertThreshold {
    pub cpu_critical: f64,
    pub cpu_warning: f64,
    pub memory_critical_percent: f64,
    pub memory_warning_percent: f64,
    pub disk_critical_percent: f64,
    pub response_time_ms_critical: u64,
}

impl Default for AlertThreshold {
    fn default() -> Self {
        Self {
            cpu_critical: 90.0,
            cpu_warning: 70.0,
            memory_critical_percent: 85.0,
            memory_warning_percent: 70.0,
            disk_critical_percent: 90.0,
            response_time_ms_critical: 5000,
        }
    }
}

#[derive(Debug, Clone)]
pub enum AlertLevel { Info, Warning, Critical, Fatal }

#[derive(Debug, Clone)]
pub struct Alert {
    pub level: AlertLevel,
    pub resource: String,
    pub message: String,
    pub value: f64,
    pub threshold: f64,
    pub timestamp: u64,
}

pub struct Sentinel {
    thresholds: AlertThreshold,
    history: Arc<Mutex<VecDeque<ResourceSnapshot>>>,
    alerts: Arc<Mutex<Vec<Alert>>>,
    max_history: usize,
}

impl Sentinel {
    pub fn new(thresholds: AlertThreshold, max_history: usize) -> Self {
        Self {
            thresholds,
            history: Arc::new(Mutex::new(VecDeque::with_capacity(max_history))),
            alerts: Arc::new(Mutex::new(Vec::new())),
            max_history,
        }
    }

    /// Record a resource snapshot and check thresholds
    pub fn record(&self, snapshot: ResourceSnapshot) -> Result<Vec<Alert>, SentinelError> {
        let mut new_alerts = Vec::new();
        let ts = snapshot.timestamp;

        // CPU check
        if snapshot.cpu_percent >= self.thresholds.cpu_critical {
            new_alerts.push(Alert {
                level: AlertLevel::Critical,
                resource: "cpu".into(),
                message: format!("CPU {}% >= {}%", snapshot.cpu_percent, self.thresholds.cpu_critical),
                value: snapshot.cpu_percent,
                threshold: self.thresholds.cpu_critical,
                timestamp: ts,
            });
        } else if snapshot.cpu_percent >= self.thresholds.cpu_warning {
            new_alerts.push(Alert {
                level: AlertLevel::Warning,
                resource: "cpu".into(),
                message: format!("CPU {}% >= {}%", snapshot.cpu_percent, self.thresholds.cpu_warning),
                value: snapshot.cpu_percent,
                threshold: self.thresholds.cpu_warning,
                timestamp: ts,
            });
        }

        // Memory check
        let mem_percent = if snapshot.memory_total_mb > 0 {
            (snapshot.memory_used_mb as f64 / snapshot.memory_total_mb as f64) * 100.0
        } else {
            0.0
        };
        if mem_percent >= self.thresholds.memory_critical_percent {
            new_alerts.push(Alert {
                level: AlertLevel::Critical,
                resource: "memory".into(),
                message: format!("Memory {:.1}% >= {}%", mem_percent, self.thresholds.memory_critical_percent),
                value: mem_percent,
                threshold: self.thresholds.memory_critical_percent,
                timestamp: ts,
            });
        } else if mem_percent >= self.thresholds.memory_warning_percent {
            new_alerts.push(Alert {
                level: AlertLevel::Warning,
                resource: "memory".into(),
                message: format!("Memory {:.1}% >= {}%", mem_percent, self.thresholds.memory_warning_percent),
                value: mem_percent,
                threshold: self.thresholds.memory_warning_percent,
                timestamp: ts,
            });
        }

        // Disk check
        if snapshot.disk_used_percent >= self.thresholds.disk_critical_percent {
            new_alerts.push(Alert {
                level: AlertLevel::Critical,
                resource: "disk".into(),
                message: format!("Disk {:.1}% >= {}%", snapshot.disk_used_percent, self.thresholds.disk_critical_percent),
                value: snapshot.disk_used_percent,
                threshold: self.thresholds.disk_critical_percent,
                timestamp: ts,
            });
        }

        // Store history
        if let Ok(mut hist) = self.history.lock() {
            if hist.len() >= self.max_history {
                hist.pop_front();
            }
            hist.push_back(snapshot);
        }

        // Store alerts
        if let Ok(mut alert_log) = self.alerts.lock() {
            for a in &new_alerts {
                alert_log.push(a.clone());
            }
        }

        Ok(new_alerts)
    }

    /// Calculate moving average of CPU over last N samples
    pub fn cpu_moving_average(&self, window: usize) -> Result<f64, SentinelError> {
        if let Ok(hist) = self.history.lock() {
            let samples: Vec<f64> = hist.iter().rev().take(window).map(|s| s.cpu_percent).collect();
            if samples.is_empty() {
                return Ok(0.0);
            }
            let sum: f64 = samples.iter().sum();
            Ok(sum / samples.len() as f64)
        } else {
            Err(SentinelError::MonitorFailed("Lock poisoned".into()))
        }
    }

    /// Detect anomaly: sudden spike compared to moving average
    pub fn detect_anomaly(&self, current_cpu: f64, window: usize) -> Result<bool, SentinelError> {
        let avg = self.cpu_moving_average(window)?;
        if avg > 0.0 {
            let deviation = (current_cpu - avg).abs() / avg;
            Ok(deviation > 0.5) // 50% deviation = anomaly
        } else {
            Ok(false)
        }
    }

    /// SLA check: uptime percentage
    pub fn calculate_uptime(&self, total_checks: u64, failed_checks: u64) -> Result<f64, SentinelError> {
        if total_checks == 0 {
            return Err(SentinelError::ConfigError("No checks recorded".into()));
        }
        let uptime = ((total_checks - failed_checks) as f64 / total_checks as f64) * 100.0;
        Ok(uptime)
    }

    /// Get current epoch timestamp
    pub fn now_epoch() -> u64 {
        SystemTime::now()
            .duration_since(SystemTime::UNIX_EPOCH)
            .map(|d| d.as_secs())
            .unwrap_or(0)
    }

    pub fn get_alerts(&self) -> Vec<Alert> {
        self.alerts.lock().map(|a| a.clone()).unwrap_or_default()
    }

    pub fn get_history(&self) -> Vec<ResourceSnapshot> {
        self.history.lock().map(|h| h.iter().cloned().collect()).unwrap_or_default()
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_cpu_critical_alert() {
        let sentinel = Sentinel::new(AlertThreshold::default(), 100);
        let snap = ResourceSnapshot {
            cpu_percent: 95.0, memory_used_mb: 4000, memory_total_mb: 8000,
            disk_used_percent: 50.0, open_connections: 100, timestamp: 1000,
        };
        let alerts = sentinel.record(snap).unwrap();
        assert!(!alerts.is_empty());
        assert!(matches!(alerts[0].level, AlertLevel::Critical));
    }

    #[test]
    fn test_moving_average() {
        let sentinel = Sentinel::new(AlertThreshold::default(), 100);
        for i in 0..10 {
            let snap = ResourceSnapshot {
                cpu_percent: 30.0 + i as f64, memory_used_mb: 2000,
                memory_total_mb: 8000, disk_used_percent: 40.0,
                open_connections: 50, timestamp: i as u64,
            };
            let _ = sentinel.record(snap);
        }
        let avg = sentinel.cpu_moving_average(5).unwrap();
        assert!(avg > 30.0 && avg < 40.0);
    }

    #[test]
    fn test_anomaly_detection() {
        let sentinel = Sentinel::new(AlertThreshold::default(), 100);
        for i in 0..20 {
            let snap = ResourceSnapshot {
                cpu_percent: 30.0, memory_used_mb: 2000,
                memory_total_mb: 8000, disk_used_percent: 40.0,
                open_connections: 50, timestamp: i,
            };
            let _ = sentinel.record(snap);
        }
        let is_anomaly = sentinel.detect_anomaly(80.0, 10).unwrap();
        assert!(is_anomaly);
    }

    #[test]
    fn test_sla_uptime() {
        let sentinel = Sentinel::new(AlertThreshold::default(), 100);
        let uptime = sentinel.calculate_uptime(1000, 1).unwrap();
        assert!(uptime > 99.8);
    }
}