16512 FIX Restart scheduler after setting "use the status of a service"

The following scenario would break scheduling

 * make host passive with `@service` -> internally sets next check to 0
 * make host active again, for example, `@up`
       -> calls `scheduleHostJob()` with `host->nextCheck()`, which is 0
       -> the invalid timestamp is rejected
       -> scheduling fails and has to be retriggered by the user

SUP-14635

Change-Id: If48579cbd371349263d19b8adcf417b112509f50

.werks/16512.md

@@ -0,0 +1,17 @@
+[//]: # (werk v2)
+# Restart scheduler after setting "use the status of a service"
+
+key        | value
+---------- | ---
+date       | 2024-06-13T14:21:01+00:00
+version    | 2.4.0b1
+class      | fix
+edition    | cee
+component  | core
+level      | 1
+compatible | yes
+
+This fixes the case where the default check scheduler would not restart after
+the user temporarily set "use the status of service" as the "host check
+command".  The corresponding log entry was "refuse to schedule invalid jobhost
+... at 1970-0101 ...".

packages/cmc/src/Core.cc

@@ -532,7 +532,12 @@ void Core::scheduleAllChecks() {
 
     for (const auto &host : _world->config_.hosts_) {
         if (host->canBeScheduled()) {
-            scheduleHostJob({host.get(), host->nextCheck()});
+            scheduleHostJob(
+                {host.get(),
+                 // check if host is currently passive
+                 host->nextCheck() == std::chrono::system_clock::time_point{}
+                     ? host->nextImmediateCheckTime()
+                     : host->nextCheck()});
         }
         for (const auto &service : Host::Services(*host)) {
             if (service->canBeScheduled()) {