synor/apps/synord/tests/stress_tests.rs

//! Stress tests for high throughput scenarios.
//!
//! These tests verify:
//! - High transaction throughput (TPS)
//! - Network under load
//! - Memory and resource management
//! - Concurrent operations
//! - Block production under pressure
//! - Large DAG handling

use std::sync::Arc;
use std::time::{Duration, Instant};

use tempfile::TempDir;
use tokio::sync::Semaphore;
use tokio::time::sleep;
use tracing::info;

use synord::config::NodeConfig;
use synord::node::{NodeState, SynorNode};

// ==================== Test Constants ====================

/// Timeout for stress test operations.
const STRESS_TIMEOUT: Duration = Duration::from_secs(60);

/// Number of concurrent operations for stress tests.
const CONCURRENT_OPS: usize = 100;

// ==================== Test Helpers ====================

/// Creates a test node configuration optimized for stress testing.
fn create_stress_config(temp_dir: &TempDir, node_index: u16) -> NodeConfig {
    let mut config = NodeConfig::for_network("devnet").unwrap();
    config.data_dir = temp_dir.path().join(format!("stress_node_{}", node_index));
    config.mining.enabled = false;

    // Use unique ports
    let port_base = 20000 + (std::process::id() % 500) as u16 * 10 + node_index * 3;
    config.p2p.listen_addr = format!("/ip4/127.0.0.1/tcp/{}", port_base);
    config.rpc.http_addr = format!("127.0.0.1:{}", port_base + 1);
    config.rpc.ws_addr = format!("127.0.0.1:{}", port_base + 2);

    // Increase limits for stress testing
    config.p2p.max_inbound = 200;
    config.p2p.max_outbound = 50;
    config.rpc.max_connections = 200;
    config.rpc.rate_limit = 0; // No rate limit for stress tests

    config
}

/// Stress test network configuration.
struct StressTestNetwork {
    nodes: Vec<Arc<SynorNode>>,
    _temp_dirs: Vec<TempDir>,
}

impl StressTestNetwork {
    async fn new(node_count: usize) -> anyhow::Result<Self> {
        let mut temp_dirs = Vec::new();
        let mut nodes = Vec::new();

        // First node (seed)
        let temp = TempDir::new()?;
        let seed_port = 20000 + (std::process::id() % 500) as u16 * 10;
        let mut config = create_stress_config(&temp, 0);
        temp_dirs.push(temp);
        nodes.push(Arc::new(SynorNode::new(config).await?));

        // Other nodes connect to seed
        for i in 1..node_count {
            let temp = TempDir::new()?;
            let mut config = create_stress_config(&temp, i as u16);
            config.p2p.seeds = vec![format!("/ip4/127.0.0.1/tcp/{}", seed_port)];
            temp_dirs.push(temp);
            nodes.push(Arc::new(SynorNode::new(config).await?));
        }

        Ok(StressTestNetwork {
            nodes,
            _temp_dirs: temp_dirs,
        })
    }

    async fn start_all(&self) -> anyhow::Result<()> {
        for node in &self.nodes {
            node.start().await?;
        }
        Ok(())
    }

    async fn stop_all(&self) -> anyhow::Result<()> {
        for node in &self.nodes {
            node.stop().await?;
        }
        Ok(())
    }
}

// ==================== Concurrent Query Tests ====================

#[tokio::test]
async fn test_concurrent_consensus_queries() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    let start = Instant::now();
    let mut handles = Vec::new();

    // Spawn concurrent consensus queries
    for i in 0..CONCURRENT_OPS {
        let node_clone = node.clone();
        let handle = tokio::spawn(async move {
            let consensus = node_clone.consensus();

            // Mix of different query types
            match i % 5 {
                0 => {
                    let _ = consensus.current_blue_score().await;
                }
                1 => {
                    let _ = consensus.current_daa_score().await;
                }
                2 => {
                    let _: Vec<[u8; 32]> = consensus.tips().await;
                }
                3 => {
                    let _ = consensus.current_height().await;
                }
                _ => {
                    let _ = consensus.current_difficulty().await;
                }
            }
        });
        handles.push(handle);
    }

    // Wait for all queries
    for handle in handles {
        handle.await.unwrap();
    }

    let elapsed = start.elapsed();
    let qps = CONCURRENT_OPS as f64 / elapsed.as_secs_f64();

    info!(
        queries = CONCURRENT_OPS,
        elapsed_ms = elapsed.as_millis(),
        qps = qps,
        "Concurrent consensus query performance"
    );

    // Should handle 100 concurrent queries quickly
    assert!(
        elapsed < Duration::from_secs(5),
        "Concurrent queries took too long"
    );

    node.stop().await.unwrap();
}

#[tokio::test]
async fn test_concurrent_network_queries() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    let start = Instant::now();
    let mut handles = Vec::new();

    for i in 0..CONCURRENT_OPS {
        let node_clone = node.clone();
        let handle = tokio::spawn(async move {
            let network = node_clone.network();

            match i % 3 {
                0 => {
                    let _ = network.peer_count().await;
                }
                1 => {
                    let _ = network.peers().await;
                }
                _ => {
                    let _ = network.stats().await;
                }
            }
        });
        handles.push(handle);
    }

    for handle in handles {
        handle.await.unwrap();
    }

    let elapsed = start.elapsed();
    info!(
        queries = CONCURRENT_OPS,
        elapsed_ms = elapsed.as_millis(),
        "Concurrent network query performance"
    );

    assert!(
        elapsed < Duration::from_secs(5),
        "Network queries took too long"
    );

    node.stop().await.unwrap();
}

// ==================== Multi-Node Stress Tests ====================

#[tokio::test]
async fn test_multi_node_concurrent_queries() {
    let network = StressTestNetwork::new(3).await.unwrap();
    network.start_all().await.unwrap();

    sleep(Duration::from_secs(2)).await;

    let start = Instant::now();
    let mut handles = Vec::new();

    // Distribute queries across all nodes
    for i in 0..CONCURRENT_OPS {
        let node = network.nodes[i % network.nodes.len()].clone();
        let handle = tokio::spawn(async move {
            let consensus = node.consensus();
            let _ = consensus.current_blue_score().await;
            let _: Vec<[u8; 32]> = consensus.tips().await;
        });
        handles.push(handle);
    }

    for handle in handles {
        handle.await.unwrap();
    }

    let elapsed = start.elapsed();
    info!(
        queries = CONCURRENT_OPS,
        nodes = network.nodes.len(),
        elapsed_ms = elapsed.as_millis(),
        "Multi-node concurrent query performance"
    );

    network.stop_all().await.unwrap();
}

#[tokio::test]
async fn test_many_node_network() {
    // Test with more nodes
    let node_count = 5;
    let network = StressTestNetwork::new(node_count).await.unwrap();
    network.start_all().await.unwrap();

    // Allow mesh to form
    sleep(Duration::from_secs(5)).await;

    // Check connectivity
    let mut total_peers = 0;
    for (i, node) in network.nodes.iter().enumerate() {
        let net = node.network();
        let peers = net.peer_count().await;
        total_peers += peers;
        info!(node = i, peers = peers, "Node peer count");
    }

    info!(
        total_peers = total_peers,
        avg_peers = total_peers / node_count,
        "Network connectivity"
    );

    // With 5 nodes, should have reasonable connectivity
    assert!(total_peers > 0, "Network should have connections");

    network.stop_all().await.unwrap();
}

// ==================== Block Announcement Flood Tests ====================

#[tokio::test]
async fn test_block_announcement_flood() {
    let network = StressTestNetwork::new(2).await.unwrap();
    network.start_all().await.unwrap();

    sleep(Duration::from_secs(2)).await;

    let start = Instant::now();

    // Flood with block announcements
    let net0 = network.nodes[0].network();
    for i in 0..100 {
        let hash = [i as u8; 32];
        net0.announce_block(hash).await;
    }

    let elapsed = start.elapsed();
    info!(
        announcements = 100,
        elapsed_ms = elapsed.as_millis(),
        rate = 100.0 / elapsed.as_secs_f64(),
        "Block announcement flood rate"
    );

    // Should handle rapid announcements
    assert!(
        elapsed < Duration::from_secs(5),
        "Block announcements too slow"
    );

    // Node should remain stable
    assert_eq!(network.nodes[0].state().await, NodeState::Running);
    assert_eq!(network.nodes[1].state().await, NodeState::Running);

    network.stop_all().await.unwrap();
}

// ==================== Memory and Resource Tests ====================

#[tokio::test]
async fn test_long_running_queries() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    // Run queries over an extended period
    let iterations = 50;
    let start = Instant::now();

    for i in 0..iterations {
        let consensus = node.consensus();
        let _ = consensus.current_blue_score().await;
        let _: Vec<[u8; 32]> = consensus.tips().await;
        let _: Vec<[u8; 32]> = consensus.get_selected_chain(100).await;

        if i % 10 == 0 {
            info!(iteration = i, "Long-running query progress");
        }
    }

    let elapsed = start.elapsed();
    info!(
        iterations = iterations,
        elapsed_ms = elapsed.as_millis(),
        avg_ms = elapsed.as_millis() / iterations as u128,
        "Long-running query performance"
    );

    // Node should remain healthy
    assert_eq!(node.state().await, NodeState::Running);

    node.stop().await.unwrap();
}

#[tokio::test]
async fn test_rapid_start_stop_cycles() {
    let temp_dir = TempDir::new().unwrap();

    for cycle in 0..5 {
        info!(cycle = cycle, "Start/stop cycle");

        let config = create_stress_config(&temp_dir, 0);
        let node = SynorNode::new(config).await.unwrap();

        node.start().await.unwrap();
        sleep(Duration::from_millis(500)).await;

        assert_eq!(node.state().await, NodeState::Running);

        node.stop().await.unwrap();
        assert_eq!(node.state().await, NodeState::Stopped);
    }

    info!("All start/stop cycles completed successfully");
}

// ==================== Sync Service Stress Tests ====================

#[tokio::test]
async fn test_concurrent_sync_queries() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    let start = Instant::now();
    let mut handles = Vec::new();

    for _ in 0..CONCURRENT_OPS {
        let node_clone = node.clone();
        let handle = tokio::spawn(async move {
            let sync = node_clone.sync();
            let _ = sync.state().await;
            let _ = sync.progress().await;
            let _ = sync.is_synced().await;
        });
        handles.push(handle);
    }

    for handle in handles {
        handle.await.unwrap();
    }

    let elapsed = start.elapsed();
    info!(
        queries = CONCURRENT_OPS * 3, // 3 calls per iteration
        elapsed_ms = elapsed.as_millis(),
        "Concurrent sync service query performance"
    );

    node.stop().await.unwrap();
}

// ==================== Mempool Stress Tests ====================

#[tokio::test]
async fn test_mempool_concurrent_access() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    let start = Instant::now();
    let mut handles = Vec::new();

    for _ in 0..CONCURRENT_OPS {
        let node_clone = node.clone();
        let handle = tokio::spawn(async move {
            let mempool = node_clone.mempool();
            let _ = mempool.size().await;
        });
        handles.push(handle);
    }

    for handle in handles {
        handle.await.unwrap();
    }

    let elapsed = start.elapsed();
    info!(
        queries = CONCURRENT_OPS,
        elapsed_ms = elapsed.as_millis(),
        "Concurrent mempool access performance"
    );

    node.stop().await.unwrap();
}

// ==================== Connection Stress Tests ====================

#[tokio::test]
async fn test_rapid_peer_operations() {
    let network = StressTestNetwork::new(2).await.unwrap();
    network.start_all().await.unwrap();

    sleep(Duration::from_secs(2)).await;

    let net0 = network.nodes[0].network();

    // Rapid peer list queries
    let start = Instant::now();
    for _ in 0..50 {
        let _ = net0.peers().await;
        let _ = net0.peer_count().await;
    }

    let elapsed = start.elapsed();
    info!(
        operations = 100,
        elapsed_ms = elapsed.as_millis(),
        "Rapid peer operations"
    );

    network.stop_all().await.unwrap();
}

#[tokio::test]
async fn test_subscription_under_load() {
    let network = StressTestNetwork::new(2).await.unwrap();
    network.start_all().await.unwrap();

    sleep(Duration::from_secs(2)).await;

    // Create multiple subscriptions
    let net0 = network.nodes[0].network();
    let net1 = network.nodes[1].network();

    let mut subscriptions = Vec::new();
    for _ in 0..10 {
        subscriptions.push(net0.subscribe());
        subscriptions.push(net1.subscribe());
    }

    // Send announcements while subscriptions exist
    for i in 0..50 {
        net0.announce_block([i as u8; 32]).await;
    }

    sleep(Duration::from_millis(500)).await;

    // Subscriptions should not cause issues
    assert_eq!(network.nodes[0].state().await, NodeState::Running);

    network.stop_all().await.unwrap();
}

// ==================== Throughput Measurement Tests ====================

#[tokio::test]
async fn test_consensus_throughput() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    let consensus = node.consensus();

    // Measure throughput of different operations
    let operations = 1000;

    // Blue score queries
    let start = Instant::now();
    for _ in 0..operations {
        let _ = consensus.current_blue_score().await;
    }
    let blue_score_elapsed = start.elapsed();

    // Tips queries
    let start = Instant::now();
    for _ in 0..operations {
        let _: Vec<[u8; 32]> = consensus.tips().await;
    }
    let tips_elapsed = start.elapsed();

    // Height queries
    let start = Instant::now();
    for _ in 0..operations {
        let _ = consensus.current_height().await;
    }
    let height_elapsed = start.elapsed();

    info!(
        blue_score_qps = operations as f64 / blue_score_elapsed.as_secs_f64(),
        tips_qps = operations as f64 / tips_elapsed.as_secs_f64(),
        height_qps = operations as f64 / height_elapsed.as_secs_f64(),
        "Consensus operation throughput"
    );

    node.stop().await.unwrap();
}

// ==================== Semaphore-Limited Stress Tests ====================

#[tokio::test]
async fn test_bounded_concurrent_operations() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    // Limit concurrency with semaphore
    let semaphore = Arc::new(Semaphore::new(50));
    let mut handles = Vec::new();

    let start = Instant::now();
    for _ in 0..500 {
        let node_clone = node.clone();
        let sem_clone = semaphore.clone();

        let handle = tokio::spawn(async move {
            let _permit = sem_clone.acquire().await.unwrap();
            let consensus = node_clone.consensus();
            let _ = consensus.current_blue_score().await;
        });
        handles.push(handle);
    }

    for handle in handles {
        handle.await.unwrap();
    }

    let elapsed = start.elapsed();
    info!(
        total_ops = 500,
        max_concurrent = 50,
        elapsed_ms = elapsed.as_millis(),
        ops_per_sec = 500.0 / elapsed.as_secs_f64(),
        "Bounded concurrent operations"
    );

    node.stop().await.unwrap();
}

// ==================== Node Info Stress Tests ====================

#[tokio::test]
async fn test_info_endpoint_stress() {
    let temp_dir = TempDir::new().unwrap();
    let config = create_stress_config(&temp_dir, 0);

    let node = Arc::new(SynorNode::new(config).await.unwrap());
    node.start().await.unwrap();

    sleep(Duration::from_secs(1)).await;

    let start = Instant::now();
    for _ in 0..100 {
        let info = node.info().await;
        // Verify info is valid
        assert!(!info.network.is_empty());
    }

    let elapsed = start.elapsed();
    info!(
        queries = 100,
        elapsed_ms = elapsed.as_millis(),
        qps = 100.0 / elapsed.as_secs_f64(),
        "Node info endpoint throughput"
    );

    node.stop().await.unwrap();
}

// ==================== Stability Under Load ====================

#[tokio::test]
async fn test_stability_under_mixed_load() {
    let network = StressTestNetwork::new(3).await.unwrap();
    network.start_all().await.unwrap();

    sleep(Duration::from_secs(2)).await;

    let start = Instant::now();
    let mut handles = Vec::new();

    // Mix of different operations
    for i in 0..150 {
        let node = network.nodes[i % network.nodes.len()].clone();

        let handle = tokio::spawn(async move {
            match i % 6 {
                0 => {
                    let consensus = node.consensus();
                    let _: Vec<[u8; 32]> = consensus.tips().await;
                }
                1 => {
                    let network_svc = node.network();
                    let _ = network_svc.peers().await;
                }
                2 => {
                    let sync = node.sync();
                    let _ = sync.progress().await;
                }
                3 => {
                    let mempool = node.mempool();
                    let _ = mempool.size().await;
                }
                4 => {
                    let _ = node.info().await;
                }
                _ => {
                    let consensus = node.consensus();
                    let _ = consensus.current_blue_score().await;
                }
            }
        });
        handles.push(handle);
    }

    for handle in handles {
        handle.await.unwrap();
    }

    let elapsed = start.elapsed();
    info!(
        operations = 150,
        elapsed_ms = elapsed.as_millis(),
        "Mixed load test completed"
    );

    // All nodes should remain running
    for (i, node) in network.nodes.iter().enumerate() {
        assert_eq!(
            node.state().await,
            NodeState::Running,
            "Node {} should be running",
            i
        );
    }

    network.stop_all().await.unwrap();
}