SOLR-17200: Fix false positive race condition in during core loading

Author: hossman

solr/CHANGES.txt

@@ -154,6 +154,8 @@ Bug Fixes
 
 * SOLR-17113: Correct how `/replication?command=details` describes errors in backup operations.  (Przemyslaw Ciezkowski via Christine Poerschke and Jason Gerlowski)
 
+* SOLR-17200: Fix false positive race condition in `/health?requireHealthyCores=true` during core loading (hossman)
+
 Dependency Upgrades
 ---------------------
 

solr/core/src/java/org/apache/solr/core/CoreContainer.java

@@ -2453,6 +2453,10 @@ public long getStatus() {
     return status;
   }
 
+  public boolean isStatusLoadComplete() {
+    return LOAD_COMPLETE == (getStatus() & LOAD_COMPLETE);
+  }
+
   public boolean hideStackTrace() {
     return cfg.hideStackTraces();
   }

solr/core/src/java/org/apache/solr/handler/admin/HealthCheckHandler.java

@@ -149,9 +149,17 @@ private void healthCheckCloudMode(SolrQueryRequest req, SolrQueryResponse rsp) {
 
     // Optionally require that all cores on this node are active if param 'requireHealthyCores=true'
     if (req.getParams().getBool(PARAM_REQUIRE_HEALTHY_CORES, false)) {
+      if (!coreContainer.isStatusLoadComplete()) {
+        rsp.add(STATUS, FAILURE);
+        rsp.setException(
+            new SolrException(
+                SolrException.ErrorCode.SERVICE_UNAVAILABLE,
+                "Host Unavailable: Core Loading not complete"));
+        return;
+      }
       Collection<CloudDescriptor> coreDescriptors =
-          coreContainer.getCores().stream()
-              .map(c -> c.getCoreDescriptor().getCloudDescriptor())
+          coreContainer.getCoreDescriptors().stream()
+              .map(cd -> cd.getCloudDescriptor())
               .collect(Collectors.toList());
       long unhealthyCores = findUnhealthyCores(coreDescriptors, clusterState);
       if (unhealthyCores > 0) {