anyllm_proxy 0.9.1

HTTP proxy translating Anthropic Messages API to OpenAI Chat Completions
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

// GET /admin/api/uptime

use crate::admin::db::{backend_history_30d, backend_uptime_pct};
use crate::admin::state::{with_db, SharedState};
use axum::{extract::State, Json};
use serde::Serialize;

/// One day in the 30-day health history calendar.
#[derive(Serialize)]
pub struct HistoryDay {
    date: String,
    /// `"up"`, `"down"`, or `"unknown"`.
    status: String,
}

/// Proxy process uptime summary.
#[derive(Serialize)]
pub struct ProxyUptimeInfo {
    /// Unix timestamp of admin server startup.
    started_at: u64,
    uptime_pct_30d: f64,
    history: Vec<HistoryDay>,
}

/// Per-backend uptime summary for the Uptime tab.
#[derive(Serialize)]
pub struct BackendUptimeInfo {
    name: String,
    status: String,
    last_checked_at: Option<i64>,
    last_latency_ms: Option<i64>,
    uptime_pct_30d: f64,
    history: Vec<HistoryDay>,
}

/// Response body for `GET /admin/api/uptime`.
#[derive(Serialize)]
pub struct UptimeResponse {
    proxy: ProxyUptimeInfo,
    backends: Vec<BackendUptimeInfo>,
}

/// GET /admin/api/uptime — returns proxy and per-backend uptime stats for the Uptime tab.
pub(super) async fn get_uptime(State(shared): State<SharedState>) -> Json<UptimeResponse> {
    let started_at = shared
        .started_at
        .duration_since(std::time::UNIX_EPOCH)
        .unwrap_or_default()
        .as_secs();

    // Snapshot backend names from RuntimeConfig (holds the admin-visible backend list).
    let backend_names: Vec<String> = {
        let cfg = shared
            .runtime_config
            .read()
            .unwrap_or_else(|e| e.into_inner());
        cfg.model_mappings.keys().cloned().collect()
    };

    let response = with_db(&shared.db, move |conn| {
        // Aggregate proxy-level history: any day where any backend was down counts as down.
        let proxy_history: Vec<HistoryDay> = conn
            .prepare(
                "SELECT date(checked_at, 'unixepoch') AS day,
                        MIN(CASE WHEN status='down' THEN 0 ELSE 1 END) AS all_up
                 FROM health_checks
                 WHERE checked_at >= strftime('%s','now') - 30*86400
                 GROUP BY day
                 ORDER BY day ASC",
            )
            .ok()
            .and_then(|mut stmt| {
                stmt.query_map([], |r| {
                    Ok(HistoryDay {
                        date: r.get(0)?,
                        status: if r.get::<_, i64>(1)? == 1 {
                            "up".to_string()
                        } else {
                            "down".to_string()
                        },
                    })
                })
                .ok()
                .map(|rows| rows.filter_map(|r| r.ok()).collect())
            })
            .unwrap_or_default();

        let backends: Vec<BackendUptimeInfo> = backend_names
            .iter()
            .map(|name| {
                let uptime_pct = backend_uptime_pct(conn, name).unwrap_or(100.0);
                let history = backend_history_30d(conn, name)
                    .unwrap_or_default()
                    .into_iter()
                    .map(|(date, status)| HistoryDay { date, status })
                    .collect();

                let (last_checked_at, last_latency_ms, current_status) = conn
                    .query_row(
                        "SELECT checked_at, latency_ms, status FROM health_checks
                         WHERE backend = ?1 ORDER BY checked_at DESC LIMIT 1",
                        [name.as_str()],
                        |r| {
                            Ok((
                                r.get::<_, i64>(0)?,
                                r.get::<_, Option<i64>>(1)?,
                                r.get::<_, String>(2)?,
                            ))
                        },
                    )
                    .map(|(ts, lat, st)| (Some(ts), lat, st))
                    .unwrap_or((None, None, "unknown".to_string()));

                BackendUptimeInfo {
                    name: name.clone(),
                    status: current_status,
                    last_checked_at,
                    last_latency_ms,
                    uptime_pct_30d: uptime_pct,
                    history,
                }
            })
            .collect();

        let down_days = proxy_history.iter().filter(|d| d.status == "down").count();
        let total_days = proxy_history.len();
        let proxy_uptime_pct = if total_days == 0 {
            100.0
        } else {
            (total_days - down_days) as f64 / total_days as f64 * 100.0
        };

        UptimeResponse {
            proxy: ProxyUptimeInfo {
                started_at,
                uptime_pct_30d: proxy_uptime_pct,
                history: proxy_history,
            },
            backends,
        }
    })
    .await;

    Json(response.unwrap_or_else(|| UptimeResponse {
        proxy: ProxyUptimeInfo {
            started_at,
            uptime_pct_30d: 100.0,
            history: vec![],
        },
        backends: vec![],
    }))
}