linkrs-server 0.1.0

Linkrs network service layer (HTTP/gRPC)
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
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184

//! Database overview handlers.
//!
//! Assembles storage and performance counters into the database overview.

use axum::{extract::State, response::Json as JsonResponse};

use crate::http::{error::HttpError, state::AppState};
use crate::storage::{
    StorageClient, StorageOperationContextOps, StorageSchemaContextOps, StorageSyncContextOps,
};
use linkrs_metrics::MetricType;
use linkrs_wire::meta::{
    CheckpointSummary, DatabaseOverviewResponse, DatabasePerformance, DatabaseSearchSummary,
    DatabaseSpaces, DatabaseStorage, LatencyPercentilesUs,
};

#[utoipa::path(
    get,
    path = "/v1/statistics/database",
    tag = "Statistics",
    responses(
        (status = 200, body = DatabaseOverviewResponse, description = "Database statistics"),
        (status = 500, description = "Internal error")
    )
)]
/// Obtain database statistics
pub async fn database<
    S: StorageClient
        + StorageSchemaContextOps
        + StorageSyncContextOps
        + StorageOperationContextOps
        + Clone
        + Send
        + Sync
        + 'static,
>(
    State(state): State<AppState<S>>,
) -> Result<JsonResponse<DatabaseOverviewResponse>, HttpError> {
    let stats_manager = state.server.get_stats_manager();
    let storage = state.server.get_storage();

    // Using `spawn_blocking` in an asynchronous context to acquire a storage lock
    let storage_stats = {
        let storage = storage.clone();
        tokio::task::spawn_blocking(move || {
            let storage = storage.read();
            storage.get_storage_stats()
        })
        .await
        .map_err(|e| HttpError::internal(format!("Failed to get storage statistics: {:?}", e)))?
    };

    Ok(JsonResponse(collect_database_overview(
        stats_manager,
        &storage_stats,
    )))
}

pub(crate) fn collect_database_overview(
    stats_manager: &linkrs_metrics::StatsManager,
    storage_stats: &crate::storage::StorageStats,
) -> DatabaseOverviewResponse {
    // Obtain statistics related to the query from both sources
    let total_queries = stats_manager.get_value(MetricType::NumQueries).unwrap_or(0);
    let active_queries = stats_manager
        .get_value(MetricType::NumActiveQueries)
        .unwrap_or(0);

    // Obtain the cache size
    let cache_size = stats_manager.query_cache_size();

    // Calculating performance metrics over the recent window.
    let recent_queries = stats_manager.get_recent_queries(100);
    let avg_latency_ms = if recent_queries.is_empty() {
        0.0
    } else {
        recent_queries
            .iter()
            .map(|q| q.total_duration_us)
            .sum::<u64>() as f64
            / recent_queries.len() as f64
            / 1000.0
    };

    // QPS from the per-second aggregation, not from recent-profile spans.
    let qps = stats_manager.get_current_qps() as f64;

    // Obtain search metrics
    let num_search_queries = stats_manager
        .get_value(MetricType::NumSearchQueries)
        .unwrap_or(0);
    let num_search_errors = stats_manager
        .get_value(MetricType::NumSearchErrors)
        .unwrap_or(0);
    let search_latency_ms = stats_manager
        .get_value(MetricType::SearchLatencyMs)
        .unwrap_or(0);
    let num_index_operations = stats_manager
        .get_value(MetricType::NumIndexOperations)
        .unwrap_or(0);
    let num_delete_operations = stats_manager
        .get_value(MetricType::NumDeleteOperations)
        .unwrap_or(0);
    let cache_hit_count = stats_manager
        .get_value(MetricType::SearchCacheHitCount)
        .unwrap_or(0);
    let cache_miss_count = stats_manager
        .get_value(MetricType::SearchCacheMissCount)
        .unwrap_or(0);

    let avg_search_latency_ms = if num_search_queries > 0 {
        search_latency_ms as f64 / num_search_queries as f64
    } else {
        0.0
    };

    let search_cache_hit_rate = if cache_hit_count + cache_miss_count > 0 {
        cache_hit_count as f64 / (cache_hit_count + cache_miss_count) as f64
    } else {
        0.0
    };

    let latency = stats_manager.query_latency_snapshot();
    let errors = stats_manager.error_snapshot();
    let storage_snapshot = stats_manager.storage_snapshot();
    let checkpoint = stats_manager.checkpoint_snapshot();

    DatabaseOverviewResponse {
        spaces: DatabaseSpaces {
            count: storage_stats.total_spaces,
            total_vertices: storage_stats.total_vertices,
            total_edges: storage_stats.total_edges,
        },
        storage: DatabaseStorage {
            // StorageStats semantics: total is allocated bytes including
            // holes and overhead, data is the live estimate, index is the
            // derived residual `total - data` (fragmentation plus overhead).
            total_size_bytes: storage_stats.total_size_bytes,
            index_size_bytes: storage_stats.index_size_bytes,
            data_size_bytes: storage_stats.data_size_bytes,
            fragmentation_permille: storage_snapshot.fragmentation_permille,
            wasted_bytes: storage_snapshot.wasted_bytes,
            tombstone_count: storage_snapshot.tombstone_count,
            tombstone_memory_bytes: storage_snapshot.tombstone_memory_bytes,
            dirty_pages: storage_snapshot.dirty_pages,
            dirty_pages_total: storage_snapshot.dirty_pages_total,
            checkpoint: CheckpointSummary {
                success_count: checkpoint.success_count,
                failure_count: checkpoint.failure_count,
                trigger_count: checkpoint.trigger_count,
                avg_duration_us: checkpoint.avg_duration_us,
                triggered_by_wal_size: checkpoint.triggered_by_wal_size,
                triggered_by_interval: checkpoint.triggered_by_interval,
                triggered_explicit: checkpoint.triggered_explicit,
            },
        },
        performance: DatabasePerformance {
            total_queries,
            active_queries,
            query_cache_size: cache_size,
            queries_per_second: qps,
            avg_latency_ms,
            cache_hit_rate: search_cache_hit_rate,
            cache_hit_rate_source: "search_cache".to_string(),
            error_total: errors.total_errors,
            latency_percentiles_us: Some(LatencyPercentilesUs {
                avg: latency.avg_us,
                p50: latency.p50_us,
                p95: latency.p95_us,
                p99: latency.p99_us,
            }),
        },
        search: DatabaseSearchSummary {
            total_queries: num_search_queries,
            total_errors: num_search_errors,
            avg_latency_ms: avg_search_latency_ms,
            total_index_operations: num_index_operations,
            total_delete_operations: num_delete_operations,
            cache_hit_count,
            cache_miss_count,
            cache_hit_rate: search_cache_hit_rate,
        },
    }
}