diff --git a/server/src/main/java/org/elasticsearch/cluster/ClusterStateTaskExecutor.java b/server/src/main/java/org/elasticsearch/cluster/ClusterStateTaskExecutor.java
index 081f8150d8c8e..fc753c2fea5a5 100644
--- a/server/src/main/java/org/elasticsearch/cluster/ClusterStateTaskExecutor.java
+++ b/server/src/main/java/org/elasticsearch/cluster/ClusterStateTaskExecutor.java
@@ -43,6 +43,10 @@ public interface ClusterStateTaskExecutor<T extends ClusterStateTaskListener> {
      */
     ClusterState execute(BatchExecutionContext<T> batchExecutionContext) throws Exception;
 
+    default boolean shouldBeThrottled() {
+        return false;
+    }
+
     /**
      * @return {@code true} iff this executor should only run on the elected master.
      */
diff --git a/server/src/main/java/org/elasticsearch/cluster/ClusterStateUpdateTask.java b/server/src/main/java/org/elasticsearch/cluster/ClusterStateUpdateTask.java
index 4f038132e7dfa..32fcd2d68a430 100644
--- a/server/src/main/java/org/elasticsearch/cluster/ClusterStateUpdateTask.java
+++ b/server/src/main/java/org/elasticsearch/cluster/ClusterStateUpdateTask.java
@@ -69,4 +69,8 @@ public final TimeValue timeout() {
     public final Priority priority() {
         return priority;
     }
+
+    public boolean shouldBeThrottled() {
+        return false;
+    }
 }
diff --git a/server/src/main/java/org/elasticsearch/cluster/service/MasterService.java b/server/src/main/java/org/elasticsearch/cluster/service/MasterService.java
index 296acc30a83f5..0ae3110e1703c 100644
--- a/server/src/main/java/org/elasticsearch/cluster/service/MasterService.java
+++ b/server/src/main/java/org/elasticsearch/cluster/service/MasterService.java
@@ -118,6 +118,11 @@ public class MasterService extends AbstractLifecycleComponent {
     private final ClusterStateUpdateStatsTracker clusterStateUpdateStatsTracker = new ClusterStateUpdateStatsTracker();
     private final StarvationWatcher starvationWatcher = new StarvationWatcher();
 
+    private static final long MAX_UPDATE_TASKS_PER_INTERVAL = 1;
+    private static final long BUDGET_INTERVAL_MILLIS = 30 * 1000;
+    private final long budgetStartTimeMillis = 0;
+    private AtomicLong budget = new AtomicLong(MAX_UPDATE_TASKS_PER_INTERVAL);
+
     public MasterService(Settings settings, ClusterSettings clusterSettings, ThreadPool threadPool, TaskManager taskManager) {
         this.nodeName = Objects.requireNonNull(Node.NODE_NAME_SETTING.get(settings));
 
@@ -227,6 +232,10 @@ private <T extends ClusterStateTaskListener> void executeAndPublishBatch(
             return;
         }
 
+        if (getTimeSince(budgetStartTimeMillis).millis() > BUDGET_INTERVAL_MILLIS) {
+            budget.set(MAX_UPDATE_TASKS_PER_INTERVAL);
+        }
+
         final long computationStartTime = threadPool.rawRelativeTimeInMillis();
         final var newClusterState = patchVersions(
             previousClusterState,
@@ -556,7 +565,7 @@ public void submitUnbatchedStateUpdateTask(String source, ClusterStateUpdateTask
         createTaskQueue("unbatched", updateTask.priority(), unbatchedExecutor).submitTask(source, updateTask, updateTask.timeout());
     }
 
-    private static class UnbatchedExecutor implements ClusterStateTaskExecutor<ClusterStateUpdateTask> {
+    private class UnbatchedExecutor implements ClusterStateTaskExecutor<ClusterStateUpdateTask> {
         @Override
         @SuppressForbidden(reason = "consuming published cluster state for legacy reasons")
         public ClusterState execute(BatchExecutionContext<ClusterStateUpdateTask> batchExecutionContext) throws Exception {
@@ -564,6 +573,9 @@ public ClusterState execute(BatchExecutionContext<ClusterStateUpdateTask> batchE
                 : "this only supports a single task but received " + batchExecutionContext.taskContexts();
             final var taskContext = batchExecutionContext.taskContexts().get(0);
             final var task = taskContext.getTask();
+            if (task.shouldBeThrottled() && budget.decrementAndGet() < 0) {
+                throw new EsRejectedExecutionException("too many cluster state updates");
+            }
             final ClusterState newState;
             try (var ignored = taskContext.captureResponseHeaders()) {
                 newState = task.execute(batchExecutionContext.initialState());
@@ -1023,7 +1035,7 @@ public String toString() {
         }
     }
 
-    private static <T extends ClusterStateTaskListener> ClusterState executeTasks(
+    private <T extends ClusterStateTaskListener> ClusterState executeTasks(
         ClusterState previousClusterState,
         List<ExecutionResult<T>> executionResults,
         ClusterStateTaskExecutor<T> executor,
@@ -1055,7 +1067,7 @@ private static <T extends ClusterStateTaskListener> boolean assertAllTasksComple
 
     static final String TEST_ONLY_EXECUTOR_MAY_CHANGE_VERSION_NUMBER_TRANSIENT_NAME = "test_only_executor_may_change_version_number";
 
-    private static <T extends ClusterStateTaskListener> ClusterState innerExecuteTasks(
+    private <T extends ClusterStateTaskListener> ClusterState innerExecuteTasks(
         ClusterState previousClusterState,
         List<ExecutionResult<T>> executionResults,
         ClusterStateTaskExecutor<T> executor,
@@ -1067,6 +1079,9 @@ private static <T extends ClusterStateTaskListener> ClusterState innerExecuteTas
             // to avoid leaking headers in production that were missed by tests
 
             try {
+                if (executor.shouldBeThrottled() && budget.decrementAndGet() < 0) {
+                    throw new EsRejectedExecutionException("too many cluster state updates");
+                }
                 final var updatedState = executor.execute(
                     new ClusterStateTaskExecutor.BatchExecutionContext<>(
                         previousClusterState,
diff --git a/server/src/test/java/org/elasticsearch/cluster/service/MasterServiceTests.java b/server/src/test/java/org/elasticsearch/cluster/service/MasterServiceTests.java
index 26faa295cf727..01af9ed8d028f 100644
--- a/server/src/test/java/org/elasticsearch/cluster/service/MasterServiceTests.java
+++ b/server/src/test/java/org/elasticsearch/cluster/service/MasterServiceTests.java
@@ -80,6 +80,7 @@
 import java.util.concurrent.atomic.AtomicReference;
 import java.util.function.UnaryOperator;
 import java.util.stream.Collectors;
+import java.util.stream.IntStream;
 
 import static java.util.Collections.emptySet;
 import static org.elasticsearch.action.support.ActionTestUtils.assertNoSuccessListener;
@@ -257,6 +258,35 @@ public void clusterStatePublished(ClusterState newClusterState) {
         assertThat(registeredActions.toString(), registeredActions, contains(MasterService.STATE_UPDATE_ACTION_NAME));
     }
 
+    public void testThrottledClusterStateUpdate() {
+        try (var master = createMasterService(true)) {
+            var failed = new AtomicInteger(0);
+            var succeeded = new AtomicInteger(0);
+            // use some listeners for this test? ref counting ...
+            IntStream.range(0, 2).forEach(i -> {
+                master.submitUnbatchedStateUpdateTask("test", new ClusterStateUpdateTask() {
+                    @Override
+                    public ClusterState execute(ClusterState currentState) throws Exception {
+                        succeeded.incrementAndGet();
+                        return currentState;
+                    }
+
+                    @Override
+                    public void onFailure(Exception e) {
+                        failed.incrementAndGet();
+                    }
+
+                    @Override
+                    public boolean shouldBeThrottled() {
+                        return true;
+                    }
+                });
+            });
+            assertThat(succeeded.get(), equalTo(1));
+            assertThat(failed.get(), equalTo(1));
+        }
+    }
+
     public void testThreadContext() throws InterruptedException {
         try (var master = createMasterService(true)) {
             final CountDownLatch latch = new CountDownLatch(1);