Write a topic's subscriber activity every one to two hours instead of every 30 to 60 minutes

Claude-Session: https://claude.ai/code/session_01DeXuhSjq5Gta5qPJgSceFe

docs/releases.md

@@ -2345,7 +2345,7 @@ and the [ntfy Android app](https://github.com/binwiederhier/ntfy-android/release
 
 **Bug fixes + maintenance:**
 
-* With `visitor-subscriber-rate-limiting` and a user database, write a UnifiedPush topic's subscriber activity about once an hour instead of every 5 minutes, spread over time, and not at all when a poll or reconnect only repeats the same subscriber; on ntfy.sh, v2.37.0's writes of ~45k topics every 5 minutes, all due in the same minute, slowed down the database ([#2151](https://github.com/binwiederhier/ntfy/pull/2151))
+* With `visitor-subscriber-rate-limiting` and a user database, write a UnifiedPush topic's subscriber activity every one to two hours instead of every 5 minutes, spread over time, and not at all when a poll or reconnect only repeats the same subscriber; on ntfy.sh, v2.37.0's writes of ~45k topics every 5 minutes, all due in the same minute, slowed down the database ([#2151](https://github.com/binwiederhier/ntfy/pull/2151))
 
 ### ntfy iOS app v1.8.0 (UNRELEASED)
 

topic/service.go

@@ -36,11 +36,11 @@ const (
 	pruneAfter    = 7 * 24 * time.Hour
 
 	// touchInterval is how often a touch moves a topic's subscriber activity at most, so an open
-	// stream, touched by the manager every minute, writes about once an hour. Each topic has its own
-	// interval between half of it and all of it (touchIntervalFor), so topics touched together,
-	// like the streams that reconnect after a restart, are not written together again. Hourly is
+	// stream, touched by the manager every minute, writes every one to two hours. Each topic has its
+	// own interval between half of it and all of it (touchIntervalFor), so topics touched together,
+	// like the streams that reconnect after a restart, are not written together again. Two hours is
 	// plenty for the rate visitor's day (rateVisitorLiveFor) and the 410's four days.
-	touchInterval = time.Hour
+	touchInterval = 2 * time.Hour
 
 	// maxRefused caps the topics known only from refused pushes, so a push flood cannot fill memory;
 	// rows read from other nodes count too, so it is effectively cluster-wide

topic/service_test.go

@@ -101,7 +101,7 @@ func TestService_State_ReturnsCopy(t *testing.T) {
 
 func TestService_Touch_Throttled(t *testing.T) {
 	// A touch moves subscriber activity at most once per the topic's own touch interval, so an open
-	// stream the manager touches every minute costs a write about once an hour only
+	// stream the manager touches every minute costs a write every one to two hours only
 	s, clock := newTestServiceWithClock(t)
 	s.SetRateVisitor("upA", "", "1.2.3.4")
 	set := clock.unix()