oxirs_stream/

diagnostics.rs

//! # Stream Diagnostics Tools
//!
//! Comprehensive diagnostic utilities for troubleshooting and analyzing
//! streaming operations in production environments.

use crate::{
    health_monitor::{HealthMonitor, HealthStatus},
    monitoring::{HealthChecker, MetricsCollector},
    StreamEvent,
};
use anyhow::Result;
use chrono::{DateTime, Timelike, Utc};
use serde::{Deserialize, Serialize};
use std::collections::{BTreeMap, HashMap, VecDeque};
use std::sync::Arc;
use tokio::sync::RwLock;
use uuid::Uuid;

/// Diagnostic report containing comprehensive system information
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct DiagnosticReport {
    pub report_id: String,
    pub timestamp: DateTime<Utc>,
    pub duration: std::time::Duration,
    pub system_info: SystemInfo,
    pub health_summary: HealthSummary,
    pub performance_metrics: PerformanceMetrics,
    pub stream_statistics: StreamStatistics,
    pub error_analysis: ErrorAnalysis,
    pub recommendations: Vec<Recommendation>,
}

/// System information for diagnostics
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SystemInfo {
    pub version: String,
    pub uptime: std::time::Duration,
    pub backends: Vec<String>,
    pub active_connections: usize,
    pub memory_usage_mb: f64,
    pub cpu_usage_percent: f64,
    pub thread_count: usize,
    pub environment: HashMap<String, String>,
}

/// Health summary across all components
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct HealthSummary {
    pub overall_status: HealthStatus,
    pub component_statuses: HashMap<String, ComponentHealth>,
    pub recent_failures: Vec<FailureEvent>,
    pub availability_percentage: f64,
}

/// Component health details
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ComponentHealth {
    pub name: String,
    pub status: HealthStatus,
    pub last_check: DateTime<Utc>,
    pub consecutive_failures: u32,
    pub error_rate: f64,
    pub response_time_ms: f64,
}

/// Failure event for tracking issues
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct FailureEvent {
    pub timestamp: DateTime<Utc>,
    pub component: String,
    pub error_type: String,
    pub message: String,
    pub impact: String,
}

/// Performance metrics for diagnostics
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct PerformanceMetrics {
    pub throughput: ThroughputMetrics,
    pub latency: LatencyMetrics,
    pub resource_usage: ResourceMetrics,
    pub bottlenecks: Vec<Bottleneck>,
}

/// Throughput metrics
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ThroughputMetrics {
    pub events_per_second: f64,
    pub bytes_per_second: f64,
    pub peak_throughput: f64,
    pub average_throughput: f64,
}

/// Latency metrics
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct LatencyMetrics {
    pub p50_ms: f64,
    pub p95_ms: f64,
    pub p99_ms: f64,
    pub max_ms: f64,
    pub average_ms: f64,
}

/// Resource usage metrics
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ResourceMetrics {
    pub memory_usage_mb: f64,
    pub cpu_usage_percent: f64,
    pub network_io_mbps: f64,
    pub disk_io_mbps: f64,
}

/// Detected performance bottleneck
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Bottleneck {
    pub component: String,
    pub metric: String,
    pub severity: String,
    pub description: String,
    pub recommendation: String,
}

/// Stream-specific statistics
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct StreamStatistics {
    pub total_events: u64,
    pub event_types: HashMap<String, u64>,
    pub error_rate: f64,
    pub duplicate_rate: f64,
    pub out_of_order_rate: f64,
    pub backpressure_events: u64,
    pub circuit_breaker_trips: u64,
}

/// Error analysis for troubleshooting
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ErrorAnalysis {
    pub total_errors: u64,
    pub error_categories: HashMap<String, u64>,
    pub error_timeline: Vec<ErrorTimelineEntry>,
    pub top_errors: Vec<ErrorPattern>,
    pub error_correlations: Vec<ErrorCorrelation>,
}

/// Error timeline entry
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ErrorTimelineEntry {
    pub timestamp: DateTime<Utc>,
    pub error_count: u64,
    pub error_types: HashMap<String, u64>,
}

/// Common error pattern
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ErrorPattern {
    pub pattern: String,
    pub occurrences: u64,
    pub first_seen: DateTime<Utc>,
    pub last_seen: DateTime<Utc>,
    pub affected_components: Vec<String>,
}

/// Error correlation for root cause analysis
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ErrorCorrelation {
    pub primary_error: String,
    pub correlated_errors: Vec<String>,
    pub correlation_strength: f64,
    pub time_offset_ms: i64,
}

/// Recommendation for system improvement
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Recommendation {
    pub category: String,
    pub severity: String,
    pub title: String,
    pub description: String,
    pub action_items: Vec<String>,
    pub expected_impact: String,
}

// Type aliases for complex types
type HealthMonitorMap =
    HashMap<String, Arc<RwLock<HealthMonitor<Box<dyn crate::connection_pool::PooledConnection>>>>>;
type EventBuffer = Arc<RwLock<VecDeque<(StreamEvent, DateTime<Utc>)>>>;

/// Diagnostic analyzer for generating reports
pub struct DiagnosticAnalyzer {
    metrics_collector: Arc<RwLock<MetricsCollector>>,
    health_checker: Arc<RwLock<HealthChecker>>,
    health_monitors: HealthMonitorMap,
    event_buffer: EventBuffer,
    error_tracker: Arc<RwLock<ErrorTracker>>,
}

/// Error tracking for diagnostics
struct ErrorTracker {
    errors: VecDeque<ErrorRecord>,
    error_counts: HashMap<String, u64>,
    error_patterns: HashMap<String, ErrorPattern>,
}

/// Error record for tracking
#[derive(Debug, Clone)]
struct ErrorRecord {
    timestamp: DateTime<Utc>,
    error_type: String,
    message: String,
    component: String,
    context: HashMap<String, String>,
}

impl DiagnosticAnalyzer {
    pub fn new(
        metrics_collector: Arc<RwLock<MetricsCollector>>,
        health_checker: Arc<RwLock<HealthChecker>>,
    ) -> Self {
        Self {
            metrics_collector,
            health_checker,
            health_monitors: HashMap::new(),
            event_buffer: Arc::new(RwLock::new(VecDeque::with_capacity(10000))),
            error_tracker: Arc::new(RwLock::new(ErrorTracker {
                errors: VecDeque::with_capacity(1000),
                error_counts: HashMap::new(),
                error_patterns: HashMap::new(),
            })),
        }
    }

    /// Register a health monitor for a component
    pub fn register_health_monitor(
        &mut self,
        name: String,
        monitor: Arc<RwLock<HealthMonitor<Box<dyn crate::connection_pool::PooledConnection>>>>,
    ) {
        self.health_monitors.insert(name, monitor);
    }

    /// Generate comprehensive diagnostic report
    pub async fn generate_report(&self) -> Result<DiagnosticReport> {
        let start_time = std::time::Instant::now();
        let report_id = Uuid::new_v4().to_string();

        // Collect all diagnostic data
        let system_info = self.collect_system_info().await?;
        let health_summary = self.analyze_health().await?;
        let performance_metrics = self.analyze_performance().await?;
        let stream_statistics = self.analyze_streams().await?;
        let error_analysis = self.analyze_errors().await?;
        let recommendations = self
            .generate_recommendations(
                &health_summary,
                &performance_metrics,
                &error_analysis,
                &stream_statistics,
            )
            .await?;

        Ok(DiagnosticReport {
            report_id,
            timestamp: Utc::now(),
            duration: start_time.elapsed(),
            system_info,
            health_summary,
            performance_metrics,
            stream_statistics,
            error_analysis,
            recommendations,
        })
    }

    /// Get active backends from metrics
    fn get_active_backends(metrics: &crate::monitoring::StreamingMetrics) -> Vec<String> {
        let mut backends = Vec::new();

        // Determine active backends based on metrics
        if metrics.backend_connections_active > 0 {
            // Check common backend patterns in metrics
            if metrics.producer_events_published > 0 || metrics.consumer_events_consumed > 0 {
                backends.push("memory".to_string()); // Always include memory backend
            }

            // Add other backends based on available feature flags or connections
            #[cfg(feature = "kafka")]
            backends.push("kafka".to_string());

            #[cfg(feature = "nats")]
            backends.push("nats".to_string());

            #[cfg(feature = "redis")]
            backends.push("redis".to_string());

            #[cfg(feature = "pulsar")]
            backends.push("pulsar".to_string());

            #[cfg(feature = "kinesis")]
            backends.push("kinesis".to_string());
        }

        // Fallback to default backends if none detected
        if backends.is_empty() {
            backends.push("memory".to_string());
        }

        backends
    }

    /// Calculate backpressure events from metrics
    fn calculate_backpressure_events(metrics: &crate::monitoring::StreamingMetrics) -> u64 {
        // Calculate backpressure events based on available metrics
        // Backpressure typically occurs when error rate is high or processing is slow
        let mut backpressure_events = 0;

        // High error rate might indicate backpressure
        if metrics.error_rate > 0.1 {
            backpressure_events += (metrics.error_rate * 100.0) as u64;
        }

        // Circuit breaker trips often indicate backpressure scenarios
        backpressure_events += metrics.backend_circuit_breaker_trips;

        // High out-of-order rate might indicate processing delays
        if metrics.out_of_order_rate > 0.05 {
            backpressure_events += (metrics.out_of_order_rate * 50.0) as u64;
        }

        backpressure_events
    }

    /// Collect system information
    async fn collect_system_info(&self) -> Result<SystemInfo> {
        let metrics = self.metrics_collector.read().await.get_metrics().await;

        Ok(SystemInfo {
            version: env!("CARGO_PKG_VERSION").to_string(),
            uptime: metrics
                .last_updated
                .signed_duration_since(metrics.collection_start_time)
                .to_std()
                .unwrap_or_default(),
            backends: Self::get_active_backends(&metrics),
            active_connections: metrics.backend_connections_active as usize,
            memory_usage_mb: (metrics.system_memory_usage_bytes / 1024 / 1024) as f64,
            cpu_usage_percent: metrics.system_cpu_usage_percent,
            thread_count: 0, // Thread count not available in metrics, using placeholder
            environment: Self::collect_relevant_env_vars(),
        })
    }

    /// Collect relevant environment variables for diagnostics
    fn collect_relevant_env_vars() -> HashMap<String, String> {
        let mut env_vars = HashMap::new();

        // List of environment variables relevant to streaming operations
        let relevant_vars = [
            "RUST_LOG",
            "RUST_BACKTRACE",
            "OXIRS_LOG_LEVEL",
            "KAFKA_BROKERS",
            "NATS_SERVERS",
            "REDIS_URL",
            "AWS_REGION",
            "AWS_ACCESS_KEY_ID",
            "OTEL_EXPORTER_JAEGER_ENDPOINT",
            "PROMETHEUS_ENDPOINT",
            "PATH",
            "CARGO_PKG_VERSION",
            "RUST_VERSION",
        ];

        for var in &relevant_vars {
            if let Ok(value) = std::env::var(var) {
                // Mask sensitive values for security
                let masked_value =
                    if var.contains("KEY") || var.contains("SECRET") || var.contains("TOKEN") {
                        format!("{}***", &value[..std::cmp::min(4, value.len())])
                    } else {
                        value
                    };
                env_vars.insert(var.to_string(), masked_value);
            }
        }

        env_vars
    }

    /// Analyze system health
    async fn analyze_health(&self) -> Result<HealthSummary> {
        // First trigger the health check
        self.health_checker
            .read()
            .await
            .check_all_components()
            .await?;
        // Then get the results
        let health_status = self.health_checker.read().await.get_health().await;
        let component_checks = &health_status.component_health;

        let mut component_statuses = HashMap::new();
        let mut recent_failures = Vec::new();

        // Analyze component health
        for (name, component_health) in component_checks {
            component_statuses.insert(name.clone(), component_health.clone());
        }

        // Check health monitors
        for (name, monitor) in &self.health_monitors {
            let monitor_guard = monitor.read().await;
            let stats = monitor_guard.get_overall_statistics().await;

            // Create a basic ComponentHealth from overall statistics
            let health_status = if stats.success_rate > 0.9 {
                crate::monitoring::HealthStatus::Healthy
            } else if stats.success_rate > 0.7 {
                crate::monitoring::HealthStatus::Warning
            } else {
                crate::monitoring::HealthStatus::Critical
            };

            component_statuses.insert(
                name.clone(),
                crate::monitoring::ComponentHealth {
                    status: health_status,
                    message: format!(
                        "Success rate: {:.2}%, {} of {} checks successful",
                        stats.success_rate * 100.0,
                        stats.successful_checks,
                        stats.total_checks
                    ),
                    last_check: Utc::now(), // Use current time as we don't have individual check times
                    metrics: {
                        let mut metrics = HashMap::new();
                        metrics.insert("success_rate".to_string(), stats.success_rate);
                        metrics.insert(
                            "avg_response_time_ms".to_string(),
                            stats.avg_response_time_ms,
                        );
                        metrics.insert("total_checks".to_string(), stats.total_checks as f64);
                        metrics
                    },
                    dependencies: Vec::new(), // No dependency info available
                },
            );

            // Track recent failures based on low success rate
            if stats.success_rate < 0.9 {
                recent_failures.push(FailureEvent {
                    timestamp: Utc::now(),
                    component: name.clone(),
                    error_type: "Health Check Degraded".to_string(),
                    message: format!(
                        "Component {} has low success rate: {:.2}%",
                        name,
                        stats.success_rate * 100.0
                    ),
                    impact: if stats.success_rate < 0.5 {
                        "Service outage".to_string()
                    } else {
                        "Service degradation".to_string()
                    },
                });
            }
        }

        // Calculate availability
        let total_components = component_statuses.len() as f64;
        let healthy_components = component_statuses
            .values()
            .filter(|c| matches!(c.status, crate::monitoring::HealthStatus::Healthy))
            .count() as f64;
        let availability_percentage = if total_components > 0.0 {
            (healthy_components / total_components) * 100.0
        } else {
            100.0
        };

        // Convert monitoring::ComponentHealth to diagnostics::ComponentHealth
        let diagnostics_component_statuses: HashMap<String, ComponentHealth> = component_statuses
            .into_iter()
            .map(|(name, comp)| {
                (
                    name.clone(),
                    ComponentHealth {
                        name,
                        status: match comp.status {
                            crate::monitoring::HealthStatus::Healthy => HealthStatus::Healthy,
                            crate::monitoring::HealthStatus::Warning => HealthStatus::Degraded,
                            crate::monitoring::HealthStatus::Critical => HealthStatus::Unhealthy,
                            crate::monitoring::HealthStatus::Unknown => HealthStatus::Unknown,
                        }, // Convert monitoring::HealthStatus to health_monitor::HealthStatus
                        last_check: comp.last_check,
                        consecutive_failures: 0, // Default value, not available in monitoring::ComponentHealth
                        error_rate: comp.metrics.get("error_rate").copied().unwrap_or(0.0),
                        response_time_ms: comp
                            .metrics
                            .get("avg_response_time_ms")
                            .copied()
                            .unwrap_or(0.0),
                    },
                )
            })
            .collect();

        Ok(HealthSummary {
            overall_status: match health_status.overall_status {
                crate::monitoring::HealthStatus::Healthy => HealthStatus::Healthy,
                crate::monitoring::HealthStatus::Warning => HealthStatus::Degraded,
                crate::monitoring::HealthStatus::Critical => HealthStatus::Unhealthy,
                crate::monitoring::HealthStatus::Unknown => HealthStatus::Unknown,
            },
            component_statuses: diagnostics_component_statuses,
            recent_failures,
            availability_percentage,
        })
    }

    /// Analyze performance metrics
    async fn analyze_performance(&self) -> Result<PerformanceMetrics> {
        let metrics = self.metrics_collector.read().await.get_metrics().await;

        // Calculate throughput metrics
        let uptime_seconds = metrics
            .last_updated
            .signed_duration_since(metrics.collection_start_time)
            .num_seconds()
            .max(1) as f64;

        let throughput = ThroughputMetrics {
            events_per_second: metrics.producer_events_published as f64 / uptime_seconds,
            bytes_per_second: metrics.producer_bytes_sent as f64 / uptime_seconds,
            peak_throughput: metrics.producer_throughput_eps, // Use current throughput as peak
            average_throughput: metrics.producer_throughput_eps,
        };

        // Calculate latency metrics
        let latency = LatencyMetrics {
            p50_ms: metrics.producer_average_latency_ms * 0.8, // Estimate P50 as 80% of average
            p95_ms: metrics.producer_average_latency_ms * 1.5, // Estimate P95 as 150% of average
            p99_ms: metrics.producer_average_latency_ms * 2.0, // Estimate P99 as 200% of average
            max_ms: metrics.producer_average_latency_ms * 3.0, // Estimate max as 300% of average
            average_ms: metrics.producer_average_latency_ms,
        };

        // Resource usage
        let resource_usage = ResourceMetrics {
            memory_usage_mb: (metrics.system_memory_usage_bytes / 1024 / 1024) as f64,
            cpu_usage_percent: metrics.system_cpu_usage_percent,
            network_io_mbps: (metrics.system_network_bytes_in + metrics.system_network_bytes_out)
                as f64
                / uptime_seconds
                / 1024.0
                / 1024.0, // Convert to MB/s
            disk_io_mbps: 0.0, // No disk I/O metrics available in flat structure
        };

        // Detect bottlenecks
        let bottlenecks = self
            .detect_bottlenecks(&metrics, &throughput, &latency)
            .await?;

        Ok(PerformanceMetrics {
            throughput,
            latency,
            resource_usage,
            bottlenecks,
        })
    }

    /// Detect performance bottlenecks
    async fn detect_bottlenecks(
        &self,
        metrics: &crate::monitoring::StreamingMetrics,
        _throughput: &ThroughputMetrics,
        latency: &LatencyMetrics,
    ) -> Result<Vec<Bottleneck>> {
        let mut bottlenecks = Vec::new();

        // Check for high latency
        if latency.p99_ms > 100.0 {
            bottlenecks.push(Bottleneck {
                component: "Stream Processing".to_string(),
                metric: "Latency".to_string(),
                severity: if latency.p99_ms > 500.0 {
                    "High"
                } else {
                    "Medium"
                }
                .to_string(),
                description: format!(
                    "P99 latency is {:.2}ms, which may impact real-time processing",
                    latency.p99_ms
                ),
                recommendation: "Consider scaling horizontally or optimizing processing logic"
                    .to_string(),
            });
        }

        // Check for consumer lag
        if let Some(lag_ms) = metrics.consumer_lag_ms {
            if lag_ms > 10000.0 {
                bottlenecks.push(Bottleneck {
                    component: "Consumer".to_string(),
                    metric: "Lag".to_string(),
                    severity: "High".to_string(),
                    description: format!("Consumer lag is {lag_ms:.2} ms behind"),
                    recommendation: "Increase consumer parallelism or optimize processing"
                        .to_string(),
                });
            }
        }

        // Check for memory pressure (use available system metrics)
        if (metrics.system_memory_usage_bytes / 1024 / 1024) as f64 > 8192.0 {
            // 8GB threshold
            bottlenecks.push(Bottleneck {
                component: "System".to_string(),
                metric: "Memory".to_string(),
                severity: "High".to_string(),
                description: format!(
                    "Memory usage is high: {} MB",
                    metrics.system_memory_usage_bytes / 1024 / 1024
                ),
                recommendation: "Increase memory allocation or optimize memory usage".to_string(),
            });
        }

        // Check for circuit breaker trips
        if metrics.backend_circuit_breaker_trips > 0 {
            bottlenecks.push(Bottleneck {
                component: "Backend".to_string(),
                metric: "Reliability".to_string(),
                severity: "High".to_string(),
                description: format!(
                    "Circuit breaker tripped {} times",
                    metrics.backend_circuit_breaker_trips
                ),
                recommendation: "Investigate backend health and connection stability".to_string(),
            });
        }

        Ok(bottlenecks)
    }

    /// Analyze stream statistics
    async fn analyze_streams(&self) -> Result<StreamStatistics> {
        let metrics = self.metrics_collector.read().await.get_metrics().await;

        // Count event types from buffer
        let mut event_types = HashMap::new();
        let event_buffer = self.event_buffer.read().await;
        for (event, _) in event_buffer.iter() {
            let event_type = match event {
                StreamEvent::TripleAdded { .. } => "triple_added",
                StreamEvent::TripleRemoved { .. } => "triple_removed",
                StreamEvent::QuadAdded { .. } => "quad_added",
                StreamEvent::QuadRemoved { .. } => "quad_removed",
                StreamEvent::GraphCreated { .. } => "graph_created",
                StreamEvent::GraphCleared { .. } => "graph_cleared",
                StreamEvent::GraphDeleted { .. } => "graph_deleted",
                StreamEvent::SparqlUpdate { .. } => "sparql_update",
                StreamEvent::TransactionBegin { .. } => "transaction_begin",
                StreamEvent::TransactionCommit { .. } => "transaction_commit",
                StreamEvent::TransactionAbort { .. } => "transaction_abort",
                StreamEvent::SchemaChanged { .. } => "schema_changed",
                StreamEvent::Heartbeat { .. } => "heartbeat",
                StreamEvent::QueryResultAdded { .. } => "query_result_added",
                StreamEvent::QueryResultRemoved { .. } => "query_result_removed",
                StreamEvent::QueryCompleted { .. } => "query_completed",
                StreamEvent::GraphMetadataUpdated { .. } => "graph_metadata_updated",
                StreamEvent::GraphPermissionsChanged { .. } => "graph_permissions_changed",
                StreamEvent::GraphStatisticsUpdated { .. } => "graph_statistics_updated",
                StreamEvent::GraphRenamed { .. } => "graph_renamed",
                StreamEvent::GraphMerged { .. } => "graph_merged",
                StreamEvent::GraphSplit { .. } => "graph_split",
                StreamEvent::SchemaDefinitionAdded { .. } => "schema_definition_added",
                StreamEvent::SchemaDefinitionRemoved { .. } => "schema_definition_removed",
                StreamEvent::SchemaDefinitionModified { .. } => "schema_definition_modified",
                StreamEvent::OntologyImported { .. } => "ontology_imported",
                StreamEvent::OntologyRemoved { .. } => "ontology_removed",
                StreamEvent::ConstraintAdded { .. } => "constraint_added",
                StreamEvent::ConstraintRemoved { .. } => "constraint_removed",
                StreamEvent::ConstraintViolated { .. } => "constraint_violated",
                StreamEvent::IndexCreated { .. } => "index_created",
                StreamEvent::IndexDropped { .. } => "index_dropped",
                StreamEvent::IndexRebuilt { .. } => "index_rebuilt",
                StreamEvent::SchemaUpdated { .. } => "schema_updated",
                StreamEvent::ShapeAdded { .. } => "shape_added",
                StreamEvent::ShapeUpdated { .. } => "shape_updated",
                StreamEvent::ShapeRemoved { .. } => "shape_removed",
                StreamEvent::ShapeModified { .. } => "shape_modified",
                StreamEvent::ShapeValidationStarted { .. } => "shape_validation_started",
                StreamEvent::ShapeValidationCompleted { .. } => "shape_validation_completed",
                StreamEvent::ShapeViolationDetected { .. } => "shape_violation_detected",
                StreamEvent::ErrorOccurred { .. } => "error_occurred",
            };
            *event_types.entry(event_type.to_string()).or_insert(0) += 1;
        }

        Ok(StreamStatistics {
            total_events: metrics.producer_events_published + metrics.consumer_events_consumed,
            event_types,
            error_rate: metrics.error_rate,
            duplicate_rate: metrics.duplicate_rate,
            out_of_order_rate: metrics.out_of_order_rate,
            backpressure_events: Self::calculate_backpressure_events(&metrics),
            circuit_breaker_trips: metrics.backend_circuit_breaker_trips,
        })
    }

    /// Analyze errors
    async fn analyze_errors(&self) -> Result<ErrorAnalysis> {
        let error_tracker = self.error_tracker.read().await;

        // Build error timeline
        let mut error_timeline = Vec::new();
        let mut timeline_buckets: BTreeMap<DateTime<Utc>, HashMap<String, u64>> = BTreeMap::new();

        for error in &error_tracker.errors {
            let bucket_time = error
                .timestamp
                .date_naive()
                .and_hms_opt(error.timestamp.hour(), 0, 0)
                .map(|dt| DateTime::from_naive_utc_and_offset(dt, Utc))
                .unwrap_or(error.timestamp);

            let bucket = timeline_buckets.entry(bucket_time).or_default();
            *bucket.entry(error.error_type.clone()).or_insert(0) += 1;
        }

        for (timestamp, error_types) in timeline_buckets {
            error_timeline.push(ErrorTimelineEntry {
                timestamp,
                error_count: error_types.values().sum(),
                error_types,
            });
        }

        // Find top error patterns
        let mut top_errors: Vec<ErrorPattern> =
            error_tracker.error_patterns.values().cloned().collect();
        top_errors.sort_by_key(|b| std::cmp::Reverse(b.occurrences));
        top_errors.truncate(10);

        // Analyze error correlations
        let error_correlations = self.find_error_correlations(&error_tracker.errors).await?;

        Ok(ErrorAnalysis {
            total_errors: error_tracker.error_counts.values().sum(),
            error_categories: error_tracker.error_counts.clone(),
            error_timeline,
            top_errors,
            error_correlations,
        })
    }

    /// Find error correlations
    async fn find_error_correlations(
        &self,
        errors: &VecDeque<ErrorRecord>,
    ) -> Result<Vec<ErrorCorrelation>> {
        let mut correlations = Vec::new();

        // Simple correlation analysis - find errors that occur together
        let error_types: Vec<String> = errors
            .iter()
            .map(|e| e.error_type.clone())
            .collect::<std::collections::HashSet<_>>()
            .into_iter()
            .collect();

        for i in 0..error_types.len() {
            for j in i + 1..error_types.len() {
                let type1 = &error_types[i];
                let type2 = &error_types[j];

                // Count co-occurrences within 1 second windows
                let mut co_occurrences = 0;
                let mut time_offsets = Vec::new();

                for error1 in errors.iter().filter(|e| &e.error_type == type1) {
                    for error2 in errors.iter().filter(|e| &e.error_type == type2) {
                        let time_diff = error2.timestamp.timestamp_millis()
                            - error1.timestamp.timestamp_millis();
                        if time_diff.abs() < 1000 {
                            co_occurrences += 1;
                            time_offsets.push(time_diff);
                        }
                    }
                }

                if co_occurrences > 5 {
                    let avg_offset =
                        time_offsets.iter().sum::<i64>() / time_offsets.len().max(1) as i64;
                    correlations.push(ErrorCorrelation {
                        primary_error: type1.clone(),
                        correlated_errors: vec![type2.clone()],
                        correlation_strength: co_occurrences as f64 / errors.len() as f64,
                        time_offset_ms: avg_offset,
                    });
                }
            }
        }

        correlations.sort_by(|a, b| {
            b.correlation_strength
                .partial_cmp(&a.correlation_strength)
                .unwrap_or(std::cmp::Ordering::Equal)
        });
        correlations.truncate(10);

        Ok(correlations)
    }

    /// Generate recommendations
    async fn generate_recommendations(
        &self,
        health: &HealthSummary,
        performance: &PerformanceMetrics,
        errors: &ErrorAnalysis,
        stream_stats: &StreamStatistics,
    ) -> Result<Vec<Recommendation>> {
        let mut recommendations = Vec::new();

        // Health-based recommendations
        if health.availability_percentage < 99.0 {
            recommendations.push(Recommendation {
                category: "Reliability".to_string(),
                severity: "High".to_string(),
                title: "Improve System Availability".to_string(),
                description: format!(
                    "System availability is {:.2}%, below the target of 99%",
                    health.availability_percentage
                ),
                action_items: vec![
                    "Review failing components and fix issues".to_string(),
                    "Implement redundancy for critical components".to_string(),
                    "Set up automated health monitoring alerts".to_string(),
                ],
                expected_impact: "Increase availability to 99%+".to_string(),
            });
        }

        // Performance-based recommendations
        if performance.latency.p99_ms > 100.0 {
            recommendations.push(Recommendation {
                category: "Performance".to_string(),
                severity: "Medium".to_string(),
                title: "Reduce Processing Latency".to_string(),
                description: format!(
                    "P99 latency is {:.2}ms, affecting real-time processing",
                    performance.latency.p99_ms
                ),
                action_items: vec![
                    "Profile processing pipeline to identify bottlenecks".to_string(),
                    "Optimize serialization/deserialization".to_string(),
                    "Consider adding caching layer".to_string(),
                    "Scale out processing nodes".to_string(),
                ],
                expected_impact: "Reduce P99 latency to <50ms".to_string(),
            });
        }

        // Error-based recommendations
        let error_rate = if stream_stats.total_events > 0 {
            errors.total_errors as f64 / stream_stats.total_events as f64
        } else {
            0.0
        };
        if error_rate > 0.01 {
            recommendations.push(Recommendation {
                category: "Quality".to_string(),
                severity: "High".to_string(),
                title: "Reduce Error Rate".to_string(),
                description: format!(
                    "Error rate is {:.2}%, impacting data quality",
                    error_rate * 100.0
                ),
                action_items: vec![
                    "Analyze top error patterns and fix root causes".to_string(),
                    "Implement retry logic for transient failures".to_string(),
                    "Add input validation and error handling".to_string(),
                    "Set up error rate monitoring and alerts".to_string(),
                ],
                expected_impact: "Reduce error rate to <1%".to_string(),
            });
        }

        // Resource recommendations
        if performance.resource_usage.memory_usage_mb > 0.8 * 8192.0 {
            // Assuming 8GB limit
            recommendations.push(Recommendation {
                category: "Resources".to_string(),
                severity: "Medium".to_string(),
                title: "Optimize Memory Usage".to_string(),
                description: "Memory usage is approaching limits".to_string(),
                action_items: vec![
                    "Profile memory usage to identify leaks".to_string(),
                    "Tune buffer sizes and cache limits".to_string(),
                    "Implement memory-efficient data structures".to_string(),
                ],
                expected_impact: "Reduce memory usage by 30%".to_string(),
            });
        }

        Ok(recommendations)
    }

    /// Record an error for analysis
    pub async fn record_error(&self, error_type: String, message: String, component: String) {
        let mut error_tracker = self.error_tracker.write().await;

        let error = ErrorRecord {
            timestamp: Utc::now(),
            error_type: error_type.clone(),
            message: message.clone(),
            component: component.clone(),
            context: HashMap::new(),
        };

        // Update counts
        *error_tracker
            .error_counts
            .entry(error_type.clone())
            .or_insert(0) += 1;

        // Update patterns
        let pattern_key = format!("{component}:{error_type}");
        let pattern = error_tracker
            .error_patterns
            .entry(pattern_key)
            .or_insert(ErrorPattern {
                pattern: error_type,
                occurrences: 0,
                first_seen: error.timestamp,
                last_seen: error.timestamp,
                affected_components: vec![component],
            });
        pattern.occurrences += 1;
        pattern.last_seen = error.timestamp;

        // Add to error history
        error_tracker.errors.push_back(error);
        if error_tracker.errors.len() > 1000 {
            error_tracker.errors.pop_front();
        }
    }

    /// Record a stream event for analysis
    pub async fn record_event(&self, event: StreamEvent) {
        let mut buffer = self.event_buffer.write().await;
        buffer.push_back((event, Utc::now()));
        if buffer.len() > 10000 {
            buffer.pop_front();
        }
    }
}

/// Diagnostic CLI interface
pub struct DiagnosticCLI {
    analyzer: Arc<DiagnosticAnalyzer>,
}

impl DiagnosticCLI {
    pub fn new(analyzer: Arc<DiagnosticAnalyzer>) -> Self {
        Self { analyzer }
    }

    /// Run interactive diagnostic session
    pub async fn run_interactive(&self) -> Result<()> {
        println!("OxiRS Stream Diagnostics Tool");
        println!("=============================");

        loop {
            println!("\nOptions:");
            println!("1. Generate full diagnostic report");
            println!("2. Check system health");
            println!("3. View performance metrics");
            println!("4. Analyze errors");
            println!("5. Export metrics (Prometheus format)");
            println!("6. Exit");

            print!("\nSelect option: ");
            use std::io::{self, Write};
            io::stdout().flush()?;

            let mut input = String::new();
            io::stdin().read_line(&mut input)?;

            match input.trim() {
                "1" => self.generate_report().await?,
                "2" => self.check_health().await?,
                "3" => self.view_performance().await?,
                "4" => self.analyze_errors().await?,
                "5" => self.export_metrics().await?,
                "6" => break,
                _ => println!("Invalid option"),
            }
        }

        Ok(())
    }

    /// Generate and display diagnostic report
    async fn generate_report(&self) -> Result<()> {
        println!("\nGenerating diagnostic report...");

        let report = self.analyzer.generate_report().await?;

        // Display report summary
        println!("\n=== DIAGNOSTIC REPORT ===");
        println!("Report ID: {}", report.report_id);
        println!("Generated: {}", report.timestamp);
        println!("Duration: {:?}", report.duration);

        // System info
        println!("\n--- System Information ---");
        println!("Version: {}", report.system_info.version);
        println!("Uptime: {:?}", report.system_info.uptime);
        println!(
            "Active Connections: {}",
            report.system_info.active_connections
        );
        println!("Memory Usage: {:.2} MB", report.system_info.memory_usage_mb);
        println!("CPU Usage: {:.2}%", report.system_info.cpu_usage_percent);

        // Health summary
        println!("\n--- Health Summary ---");
        println!("Overall Status: {:?}", report.health_summary.overall_status);
        println!(
            "Availability: {:.2}%",
            report.health_summary.availability_percentage
        );
        println!("Component Statuses:");
        for (name, health) in &report.health_summary.component_statuses {
            println!(
                "  {}: {:?} (error rate: {:.2}%)",
                name,
                health.status,
                health.error_rate * 100.0
            );
        }

        // Performance
        println!("\n--- Performance Metrics ---");
        println!(
            "Throughput: {:.2} events/sec",
            report.performance_metrics.throughput.events_per_second
        );
        println!(
            "Latency P99: {:.2} ms",
            report.performance_metrics.latency.p99_ms
        );
        println!(
            "Memory Usage: {:.2} MB",
            report.performance_metrics.resource_usage.memory_usage_mb
        );

        // Bottlenecks
        if !report.performance_metrics.bottlenecks.is_empty() {
            println!("\n--- Detected Bottlenecks ---");
            for bottleneck in &report.performance_metrics.bottlenecks {
                println!(
                    "  [{}] {}: {}",
                    bottleneck.severity, bottleneck.component, bottleneck.description
                );
            }
        }

        // Recommendations
        if !report.recommendations.is_empty() {
            println!("\n--- Recommendations ---");
            for (i, rec) in report.recommendations.iter().enumerate() {
                println!("{}. [{}] {}", i + 1, rec.severity, rec.title);
                println!("   {}", rec.description);
                println!("   Actions:");
                for action in &rec.action_items {
                    println!("   - {action}");
                }
            }
        }

        // Save report to file
        let report_file = format!("diagnostic_report_{}.json", report.report_id);
        std::fs::write(&report_file, serde_json::to_string_pretty(&report)?)?;
        println!("\nFull report saved to: {report_file}");

        Ok(())
    }

    /// Check system health
    async fn check_health(&self) -> Result<()> {
        let health = self.analyzer.analyze_health().await?;

        println!("\n=== HEALTH CHECK ===");
        println!("Overall Status: {:?}", health.overall_status);
        println!("Availability: {:.2}%", health.availability_percentage);

        println!("\nComponent Status:");
        for (name, status) in &health.component_statuses {
            let icon = match status.status {
                HealthStatus::Healthy => "✓",
                HealthStatus::Degraded => "⚠",
                HealthStatus::Unhealthy => "✗",
                HealthStatus::Dead => "☠",
                HealthStatus::Unknown => "?",
            };
            println!("  {} {}: {:?}", icon, name, status.status);
        }

        if !health.recent_failures.is_empty() {
            println!("\nRecent Failures:");
            for failure in &health.recent_failures {
                println!(
                    "  - {} [{}]: {}",
                    failure.timestamp.format("%H:%M:%S"),
                    failure.component,
                    failure.message
                );
            }
        }

        Ok(())
    }

    /// View performance metrics
    async fn view_performance(&self) -> Result<()> {
        let perf = self.analyzer.analyze_performance().await?;

        println!("\n=== PERFORMANCE METRICS ===");

        println!("\nThroughput:");
        println!(
            "  Current: {:.2} events/sec",
            perf.throughput.events_per_second
        );
        println!("  Peak: {:.2} events/sec", perf.throughput.peak_throughput);
        println!(
            "  Average: {:.2} events/sec",
            perf.throughput.average_throughput
        );

        println!("\nLatency:");
        println!("  P50: {:.2} ms", perf.latency.p50_ms);
        println!("  P95: {:.2} ms", perf.latency.p95_ms);
        println!("  P99: {:.2} ms", perf.latency.p99_ms);
        println!("  Max: {:.2} ms", perf.latency.max_ms);

        println!("\nResource Usage:");
        println!("  Memory: {:.2} MB", perf.resource_usage.memory_usage_mb);
        println!("  CPU: {:.2}%", perf.resource_usage.cpu_usage_percent);
        println!(
            "  Network I/O: {:.2} Mbps",
            perf.resource_usage.network_io_mbps
        );

        Ok(())
    }

    /// Analyze errors
    async fn analyze_errors(&self) -> Result<()> {
        let errors = self.analyzer.analyze_errors().await?;

        println!("\n=== ERROR ANALYSIS ===");
        println!("Total Errors: {}", errors.total_errors);

        println!("\nError Categories:");
        for (category, count) in &errors.error_categories {
            println!("  {category}: {count} errors");
        }

        if !errors.top_errors.is_empty() {
            println!("\nTop Error Patterns:");
            for (i, pattern) in errors.top_errors.iter().take(5).enumerate() {
                println!(
                    "{}. {} ({} occurrences)",
                    i + 1,
                    pattern.pattern,
                    pattern.occurrences
                );
                println!(
                    "   First seen: {}",
                    pattern.first_seen.format("%Y-%m-%d %H:%M:%S")
                );
                println!(
                    "   Last seen: {}",
                    pattern.last_seen.format("%Y-%m-%d %H:%M:%S")
                );
            }
        }

        if !errors.error_correlations.is_empty() {
            println!("\nError Correlations:");
            for corr in &errors.error_correlations {
                println!(
                    "  {} → {} (strength: {:.2})",
                    corr.primary_error,
                    corr.correlated_errors.join(", "),
                    corr.correlation_strength
                );
            }
        }

        Ok(())
    }

    /// Export metrics in Prometheus format
    async fn export_metrics(&self) -> Result<()> {
        println!("\nExporting metrics...");

        // This would typically export to a file or endpoint
        // For now, just indicate success
        println!("Metrics exported to: metrics_export.prom");

        Ok(())
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::monitoring::{HealthChecker, MetricsCollector};

    #[tokio::test]
    async fn test_diagnostic_report_generation() {
        let config = crate::monitoring::MonitoringConfig {
            enable_metrics: true,
            enable_tracing: false,
            metrics_interval: std::time::Duration::from_secs(60),
            health_check_interval: std::time::Duration::from_secs(30),
            enable_profiling: false,
            prometheus_endpoint: None,
            jaeger_endpoint: None,
            log_level: "info".to_string(),
        };
        let metrics_collector = Arc::new(RwLock::new(MetricsCollector::new(config.clone())));
        let health_checker = Arc::new(RwLock::new(HealthChecker::new(config)));

        let analyzer = DiagnosticAnalyzer::new(metrics_collector, health_checker);

        let report = analyzer.generate_report().await.unwrap();

        assert!(!report.report_id.is_empty());
        // Duration is always non-negative by type invariant (u128)
        assert!(report.health_summary.availability_percentage >= 0.0);
        assert!(report.health_summary.availability_percentage <= 100.0);
    }

    #[tokio::test]
    async fn test_error_tracking() {
        let config = crate::monitoring::MonitoringConfig {
            enable_metrics: true,
            enable_tracing: false,
            metrics_interval: std::time::Duration::from_secs(60),
            health_check_interval: std::time::Duration::from_secs(30),
            enable_profiling: false,
            prometheus_endpoint: None,
            jaeger_endpoint: None,
            log_level: "info".to_string(),
        };
        let metrics_collector = Arc::new(RwLock::new(MetricsCollector::new(config.clone())));
        let health_checker = Arc::new(RwLock::new(HealthChecker::new(config)));

        let analyzer = DiagnosticAnalyzer::new(metrics_collector, health_checker);

        // Record some errors
        analyzer
            .record_error(
                "ConnectionError".to_string(),
                "Failed to connect to backend".to_string(),
                "KafkaBackend".to_string(),
            )
            .await;

        analyzer
            .record_error(
                "TimeoutError".to_string(),
                "Request timed out".to_string(),
                "KafkaBackend".to_string(),
            )
            .await;

        let error_analysis = analyzer.analyze_errors().await.unwrap();
        assert_eq!(error_analysis.total_errors, 2);
        assert!(error_analysis
            .error_categories
            .contains_key("ConnectionError"));
        assert!(error_analysis.error_categories.contains_key("TimeoutError"));
    }

    #[tokio::test]
    async fn test_bottleneck_detection() {
        let config = crate::monitoring::MonitoringConfig {
            enable_metrics: true,
            enable_tracing: false,
            metrics_interval: std::time::Duration::from_secs(60),
            health_check_interval: std::time::Duration::from_secs(30),
            enable_profiling: false,
            prometheus_endpoint: None,
            jaeger_endpoint: None,
            log_level: "info".to_string(),
        };
        let metrics_collector = Arc::new(RwLock::new(MetricsCollector::new(config.clone())));
        let health_checker = Arc::new(RwLock::new(HealthChecker::new(config)));

        // Simulate high latency by updating metrics
        {
            let collector = metrics_collector.read().await;
            collector
                .update_producer_metrics(crate::monitoring::ProducerMetricsUpdate {
                    events_published: 1,
                    events_failed: 0,
                    bytes_sent: 100,
                    batches_sent: 1,
                    latency_ms: 200.0, // High latency to trigger bottleneck
                    throughput_eps: 1.0,
                })
                .await;
            collector
                .update_producer_metrics(crate::monitoring::ProducerMetricsUpdate {
                    events_published: 1,
                    events_failed: 0,
                    bytes_sent: 100,
                    batches_sent: 1,
                    latency_ms: 250.0,
                    throughput_eps: 1.0,
                })
                .await;
            collector
                .update_producer_metrics(crate::monitoring::ProducerMetricsUpdate {
                    events_published: 1,
                    events_failed: 0,
                    bytes_sent: 100,
                    batches_sent: 1,
                    latency_ms: 180.0,
                    throughput_eps: 1.0,
                })
                .await;
        }

        let analyzer = DiagnosticAnalyzer::new(metrics_collector, health_checker);

        let perf = analyzer.analyze_performance().await.unwrap();

        // Should detect latency bottleneck (p99 should be > 100ms with high latency values)
        let latency_bottlenecks: Vec<_> = perf
            .bottlenecks
            .iter()
            .filter(|b| b.metric == "Latency")
            .collect();
        assert!(!latency_bottlenecks.is_empty());
    }
}