feat(uptime): Extract domain and suffix from urls when creating uptim…

…e monitors

This uses `tldextract` to extract the domain and suffix from urls, for use in limiting number of monitors per domain, analytics and detecting anomalous failures across monitors.

src/sentry/uptime/subscriptions/subscriptions.py

@@ -3,6 +3,7 @@
 
 from sentry.models.project import Project
 from sentry.types.actor import Actor
+from sentry.uptime.detectors.url_extraction import extractor
 from sentry.uptime.models import (
     ProjectUptimeSubscription,
     ProjectUptimeSubscriptionMode,
@@ -28,13 +29,20 @@ def create_uptime_subscription(
     Creates a new uptime subscription. This creates the row in postgres, and fires a task that will send the config
     to the uptime check system.
     """
+    # We extract the domain and suffix of the url here. This is used to prevent there being too many checks to a single
+    # domain.
+    # We enable private PSL domains so that hosting services that use subdomains are treated as suffixes for the
+    # purposes of monitoring.
+    result = extractor.extract_str(url, include_psl_private_domains=True)
     subscription, created = UptimeSubscription.objects.get_or_create(
         url=url,
         interval_seconds=interval_seconds,
         defaults={
             "status": UptimeSubscription.Status.CREATING.value,
             "type": UPTIME_SUBSCRIPTION_TYPE,
             "timeout_ms": timeout_ms,
+            "url_domain": result.domain,
+            "url_suffix": result.suffix,
         },
     )
     if subscription.status == UptimeSubscription.Status.DELETING.value:

tests/sentry/uptime/subscriptions/test_subscriptions.py

@@ -35,6 +35,24 @@ def test(self):
         assert uptime_sub.status == UptimeSubscription.Status.ACTIVE.value
         assert uptime_sub.type == UPTIME_SUBSCRIPTION_TYPE
         assert uptime_sub.url == url
+        assert uptime_sub.url_domain == "sentry"
+        assert uptime_sub.url_suffix == "io"
+        assert uptime_sub.interval_seconds == uptime_sub.interval_seconds
+        assert uptime_sub.timeout_ms == timeout_ms
+
+    def test_private_domain_suffix(self):
+        url = "https://test.vercel.app"
+        interval_seconds = 300
+        timeout_ms = 500
+        uptime_sub = create_uptime_subscription(url, interval_seconds, timeout_ms)
+        # Subscription.subscription_id ends up set in the task, so refresh
+        uptime_sub.refresh_from_db()
+        assert uptime_sub.subscription_id is None
+        assert uptime_sub.status == UptimeSubscription.Status.CREATING.value
+        assert uptime_sub.type == UPTIME_SUBSCRIPTION_TYPE
+        assert uptime_sub.url == url
+        assert uptime_sub.url_domain == "test"
+        assert uptime_sub.url_suffix == "vercel.app"
         assert uptime_sub.interval_seconds == uptime_sub.interval_seconds
         assert uptime_sub.timeout_ms == timeout_ms
 
@@ -81,6 +99,8 @@ def test_without_task(self):
         assert uptime_sub.status == UptimeSubscription.Status.CREATING.value
         assert uptime_sub.type == UPTIME_SUBSCRIPTION_TYPE
         assert uptime_sub.url == url
+        assert uptime_sub.url_domain == "sentry"
+        assert uptime_sub.url_suffix == "io"
         assert uptime_sub.interval_seconds == uptime_sub.interval_seconds
         assert uptime_sub.timeout_ms == timeout_ms