fix(uptime): log when we detect missed uptime checkins (#91733)

If a check result comes in that (according to schedule) is much later
than when we expected our next check, record the number of missed checks
in logs and datadog.

If we somehow get an intervening check, we'll just drop it as per
preceding logic.

Author: klochek

src/sentry/uptime/consumers/results_consumer.py

@@ -464,6 +464,35 @@ def handle_result(self, subscription: UptimeSubscription | None, result: CheckRe
             )
             return
 
+        subscription_interval_ms = 1000 * subscription.interval_seconds
+        num_intervals = (
+            result["scheduled_check_time_ms"] - last_update_ms
+        ) / subscription_interval_ms
+
+        # If the scheduled check is two or more intervals since the last seen check, we can declare the
+        # intervening checks missed.
+        if last_update_raw is not None and num_intervals > 1:
+            num_missed_checks = num_intervals - 1
+            metrics.distribution(
+                "uptime.result_processer.num_missing_check",
+                num_missed_checks,
+                tags=metric_tags,
+            )
+            logger.info(
+                "uptime.result_processor.num_missing_check",
+                extra={"num_missed_checks": num_missed_checks, **result},
+            )
+            if num_intervals != int(num_intervals):
+                logger.info(
+                    "uptime.result_processor.invalid_check_interval",
+                    extra={
+                        "last_update_ms": last_update_ms,
+                        "current_update_ms": result["scheduled_check_time_ms"],
+                        "interval_ms": subscription_interval_ms,
+                        **result,
+                    },
+                )
+
         if features.has("organizations:uptime-detailed-logging", organization):
             logger.info("handle_result_for_project.after_dedupe", extra=result)