bijux-dag-runtime 0.4.0

Execution engine, replay semantics, and runtime policy layer for Bijux DAG graphs.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129

use crate::{NodeStatus, RunMetrics, RuntimeConfig, SchedulerMetrics};
use bijux_dag_artifacts::{FailurePropagationRecord, NodeCounts};
use serde_json::Value;

pub fn build_run_metrics(
    node_counts: &NodeCounts,
    graph_node_count: usize,
    options: &RuntimeConfig,
    finished_unix_ms: u128,
    started_unix_ms: u128,
    cache_hits: u64,
    output_count: usize,
) -> RunMetrics {
    let total_nodes = graph_node_count.max(1) as f64;
    RunMetrics {
        makespan_ms: finished_unix_ms.saturating_sub(started_unix_ms),
        success_ratio: node_counts.success as f64 / total_nodes,
        parallelism_utilization: (node_counts.success + node_counts.cached) as f64
            / (options.jobs.max(1) as f64 * total_nodes).max(1.0),
        cache_reuse_ratio: cache_hits as f64 / total_nodes,
        artifact_volume_bytes: output_count as u64,
        planning_ms: 0,
        scheduling_wait_ms: 0,
        execution_ms: finished_unix_ms.saturating_sub(started_unix_ms),
        trace_write_ms: 0,
        manifest_finalize_ms: 0,
        replay_compare_ms: 0,
    }
}

pub fn build_scheduler_metrics(
    node_counts: &NodeCounts,
    run_log_index: &[Value],
    options: &RuntimeConfig,
    failure_propagation_records: &[FailurePropagationRecord],
) -> SchedulerMetrics {
    SchedulerMetrics {
        queue_depth: 0,
        ready_count: 0,
        running_count: 0,
        completed_count: (node_counts.success
            + node_counts.failed
            + node_counts.skipped
            + node_counts.cached) as usize,
        retry_count: count_runtime_events(run_log_index, "node_attempt_started"),
        cache_hit_count: count_runtime_events(run_log_index, "cache_hit"),
        cache_miss_count: count_runtime_events(run_log_index, "cache_miss"),
        failure_count: count_runtime_events(run_log_index, "node_failed"),
        starvation_count: failure_propagation_records
            .iter()
            .filter(|record| record.reason == "budget")
            .count() as u64,
        dispatch_latency_ms: 0,
        concurrency_pressure: (options.jobs.max(1) as f64)
            / (options.scheduler_policy.max_parallelism.max(1) as f64),
    }
}

fn count_runtime_events(run_log_index: &[Value], name: &str) -> u64 {
    run_log_index
        .iter()
        .filter(|row| row.get("event").and_then(|v| v.as_str()) == Some(name))
        .count() as u64
}

pub fn count_cache_hits(status_map: &std::collections::HashMap<String, NodeStatus>) -> u64 {
    status_map.values().filter(|status| matches!(status, NodeStatus::Cached)).count() as u64
}

#[cfg(test)]
mod tests {
    use super::{build_run_metrics, build_scheduler_metrics, count_cache_hits};
    use crate::{NodeStatus, RuntimeConfig};
    use bijux_dag_artifacts::{FailurePropagationRecord, NodeCounts};
    use serde_json::json;
    use std::collections::HashMap;

    #[test]
    fn run_metrics_shape_is_stable_for_finished_run() {
        let counts = NodeCounts { success: 3, failed: 1, skipped: 0, cached: 2, cancelled: 0 };
        let metrics = build_run_metrics(&counts, 6, &RuntimeConfig::default(), 1_500, 1_000, 2, 4);
        assert_eq!(metrics.makespan_ms, 500);
        assert_eq!(metrics.execution_ms, 500);
        assert!(metrics.success_ratio > 0.0);
        assert!(metrics.cache_reuse_ratio > 0.0);
    }

    #[test]
    fn scheduler_metrics_counts_events_and_budget_starvation() {
        let counts = NodeCounts { success: 2, failed: 1, skipped: 1, cached: 1, cancelled: 0 };
        let log = vec![
            json!({"event":"node_attempt_started"}),
            json!({"event":"cache_hit"}),
            json!({"event":"cache_miss"}),
            json!({"event":"node_failed"}),
        ];
        let failures = vec![
            FailurePropagationRecord {
                node_id: "blocked".to_string(),
                status: "skipped".to_string(),
                reason: "budget".to_string(),
                propagation_mode: None,
                blocking_nodes: Vec::new(),
            },
            FailurePropagationRecord {
                node_id: "other".to_string(),
                status: "failed".to_string(),
                reason: "other".to_string(),
                propagation_mode: None,
                blocking_nodes: Vec::new(),
            },
        ];
        let metrics = build_scheduler_metrics(&counts, &log, &RuntimeConfig::default(), &failures);
        assert_eq!(metrics.retry_count, 1);
        assert_eq!(metrics.cache_hit_count, 1);
        assert_eq!(metrics.cache_miss_count, 1);
        assert_eq!(metrics.failure_count, 1);
        assert_eq!(metrics.starvation_count, 1);
    }

    #[test]
    fn cache_hit_counter_tracks_only_cached_nodes() {
        let mut map = HashMap::new();
        map.insert("a".to_string(), NodeStatus::Cached);
        map.insert("b".to_string(), NodeStatus::Success);
        map.insert("c".to_string(), NodeStatus::Cached);
        assert_eq!(count_cache_hits(&map), 2);
    }
}