Describe STALE_STATE_CONFIG in ClusterFormationFH (#53878)

We mark cluster states persisted on master-ineligible nodes as
potentially-stale using the voting configuration `{STALE_STATE_CONFIG}` which
prevents these nodes from being elected as master if they are restarted as
master-eligible. Today we do not handle this special voting configuration
differently in the `ClusterFormationFailureHandler`, leading to a mysterious
message `an election requires a node with id [STALE_STATE_CONFIG]` if the
election does not succeed.

This commit adds a special case description for this situation to explain
better why this node cannot win an election.

Closes #53734

server/src/main/java/org/elasticsearch/cluster/coordination/ClusterFormationFailureHelper.java

@@ -30,6 +30,7 @@
 import org.elasticsearch.common.transport.TransportAddress;
 import org.elasticsearch.common.unit.TimeValue;
 import org.elasticsearch.common.util.concurrent.AbstractRunnable;
+import org.elasticsearch.gateway.GatewayMetaState;
 import org.elasticsearch.threadpool.ThreadPool;
 import org.elasticsearch.threadpool.ThreadPool.Names;
 
@@ -206,7 +207,12 @@ private String describeQuorum(VotingConfiguration votingConfiguration) {
             assert requiredNodes <= realNodeIds.size() : nodeIds;
 
             if (nodeIds.size() == 1) {
-                return "a node with id " + realNodeIds;
+                if (nodeIds.contains(GatewayMetaState.STALE_STATE_CONFIG_NODE_ID)) {
+                    return "one or more nodes that have already participated as master-eligible nodes in the cluster but this node was " +
+                        "not master-eligible the last time it joined the cluster";
+                } else {
+                    return "a node with id " + realNodeIds;
+                }
             } else if (nodeIds.size() == 2) {
                 return "two nodes with ids " + realNodeIds;
             } else {

server/src/main/java/org/elasticsearch/gateway/GatewayMetaState.java

@@ -78,6 +78,13 @@
  */
 public class GatewayMetaState implements Closeable {
 
+    /**
+     * Fake node ID for a voting configuration written by a master-ineligible data node to indicate that its on-disk state is potentially
+     * stale (since it is written asynchronously after application, rather than before acceptance). This node ID means that if the node is
+     * restarted as a master-eligible node then it does not win any elections until it has received a fresh cluster state.
+     */
+    public static final String STALE_STATE_CONFIG_NODE_ID = "STALE_STATE_CONFIG";
+
     // Set by calling start()
     private final SetOnce<PersistedState> persistedState = new SetOnce<>();
 
@@ -360,7 +367,7 @@ protected void doRun() {
         }
 
         static final CoordinationMetaData.VotingConfiguration staleStateConfiguration =
-            new CoordinationMetaData.VotingConfiguration(Collections.singleton("STALE_STATE_CONFIG"));
+            new CoordinationMetaData.VotingConfiguration(Collections.singleton(STALE_STATE_CONFIG_NODE_ID));
 
         static ClusterState resetVotingConfiguration(ClusterState clusterState) {
             CoordinationMetaData newCoordinationMetaData = CoordinationMetaData.builder(clusterState.coordinationMetaData())

server/src/test/java/org/elasticsearch/cluster/coordination/ClusterFormationFailureHelperTests.java

@@ -29,6 +29,7 @@
 import org.elasticsearch.cluster.node.DiscoveryNodes;
 import org.elasticsearch.common.settings.Settings;
 import org.elasticsearch.common.transport.TransportAddress;
+import org.elasticsearch.gateway.GatewayMetaState;
 import org.elasticsearch.test.ESTestCase;
 
 import java.util.Arrays;
@@ -395,5 +396,14 @@ public void testDescriptionAfterBootstrapping() {
                     "have discovered [] which is not a quorum; " +
                     "discovery will continue using [] from hosts providers and [" + otherMasterNode + ", " + localNode +
                     "] from last-known cluster state; node term 0, last-accepted version 0 in term 0")));
+
+        assertThat(new ClusterFormationState(Settings.EMPTY, state(localNode, GatewayMetaState.STALE_STATE_CONFIG_NODE_ID), emptyList(),
+                emptyList(), 0L, electionStrategy).getDescription(),
+            is("master not discovered or elected yet, an election requires one or more nodes that have already participated as " +
+                "master-eligible nodes in the cluster but this node was not master-eligible the last time it joined the cluster, " +
+                "have discovered [] which is not a quorum; " +
+                "discovery will continue using [] from hosts providers and [" + localNode +
+                "] from last-known cluster state; node term 0, last-accepted version 0 in term 0"));
+
     }
 }