feat: per-region due scheduling + run_id to group pings across regions

apps/monitor/Cargo.toml

@@ -20,3 +20,4 @@ webpki-roots = "0.26"
 x509-parser  = "0.16"
 tokio-rustls = "0.26"
 chrono       = { version = "0.4", features = ["serde"] }
+uuid         = { version = "1", features = ["v4"] }

apps/monitor/src/runner.rs

@@ -33,6 +33,9 @@ pub async fn fetch_and_run(
     let n = monitors.len();
     if n == 0 { return Ok(0); }
 
+    // One run_id per fetch batch — groups all checks from this cycle across regions
+    let run_id = uuid::Uuid::new_v4().to_string();
+
     // Spawn all checks — fire and forget, skip if already in-flight
     let mut spawned = 0usize;
     for monitor in monitors {
@@ -49,12 +52,13 @@ pub async fn fetch_and_run(
         let coordinator_url = coordinator_url.to_string();
         let token = token.to_string();
         let region_owned = region.to_string();
+        let run_id_owned = run_id.clone();
         let in_flight = in_flight.clone();
         tokio::spawn(async move {
             let timeout_ms = monitor.timeout_ms.unwrap_or(30000);
             // Hard deadline: timeout + 5s buffer, so hung checks always resolve
             let deadline = std::time::Duration::from_millis(timeout_ms + 5000);
-            let result = match tokio::time::timeout(deadline, run_check(&client, &monitor, monitor.scheduled_at.clone(), &region_owned)).await {
+            let result = match tokio::time::timeout(deadline, run_check(&client, &monitor, monitor.scheduled_at.clone(), &region_owned, &run_id_owned)).await {
                 Ok(r) => r,
                 Err(_) => PingResult {
                     monitor_id: monitor.id.clone(),
@@ -67,6 +71,7 @@ pub async fn fetch_and_run(
                     cert_expiry_days: None,
                     meta: None,
                     region: if region_owned.is_empty() { None } else { Some(region_owned.clone()) },
+                    run_id: Some(run_id_owned.clone()),
                 },
             };
             // Post result first, then clear in-flight — this prevents the next
@@ -81,7 +86,7 @@ pub async fn fetch_and_run(
     Ok(spawned)
 }
 
-async fn run_check(client: &reqwest::Client, monitor: &Monitor, scheduled_at: Option<String>, region: &str) -> PingResult {
+async fn run_check(client: &reqwest::Client, monitor: &Monitor, scheduled_at: Option<String>, region: &str, run_id: &str) -> PingResult {
     // Compute jitter: how late we actually started vs when we were scheduled
     let jitter_ms: Option<i64> = scheduled_at.as_deref().and_then(|s| {
         let scheduled = chrono::DateTime::parse_from_rfc3339(s).ok()?;
@@ -135,6 +140,7 @@ async fn run_check(client: &reqwest::Client, monitor: &Monitor, scheduled_at: Op
                 cert_expiry_days: None,
                 meta: None,
                 region: if region.is_empty() { None } else { Some(region.to_string()) },
+                run_id: Some(run_id.to_string()),
             }
         },
         Ok((status_code, headers, body)) => {
@@ -195,6 +201,7 @@ async fn run_check(client: &reqwest::Client, monitor: &Monitor, scheduled_at: Op
                 cert_expiry_days,
                 meta: Some(meta),
                 region: if region.is_empty() { None } else { Some(region.to_string()) },
+                run_id: Some(run_id.to_string()),
             }
         }
     }

apps/monitor/src/types.rs

@@ -28,4 +28,5 @@ pub struct PingResult {
     pub cert_expiry_days: Option<i64>,
     pub meta:             Option<Value>,
     pub region:           Option<String>,
+    pub run_id:           Option<String>,
 }