FullClusterRestartIT.testRecovery should wait for all initializing shards

Shards that are not fully bake may not execute the flush call, causing
subsequent unexpected translog recoveries which fails the test.

Closes #27817

Author: bleskes

qa/full-cluster-restart/src/test/java/org/elasticsearch/upgrades/FullClusterRestartIT.java

@@ -710,6 +710,9 @@ public void testRecovery() throws IOException {
             shouldHaveTranslog = randomBoolean();
 
             indexRandomDocuments(count, true, true, i -> jsonBuilder().startObject().field("field", "value").endObject());
+
+            // make sure all recoveries are done
+            ensureNoInitializingShards();
             // Explicitly flush so we're sure to have a bunch of documents in the Lucene index
             client().performRequest("POST", "/_flush");
             if (shouldHaveTranslog) {