Move Get Snapshots Serialization to Management Pool (#83215)

It's in the title. For large index counts generating the response,
both for transport as well as REST layer gets quite expensive.
Better generate it off the transport threads.

docs/changelog/83215.yaml

@@ -0,0 +1,5 @@
+pr: 83215
+summary: Move Get Snapshots Serialization to Management Pool
+area: Snapshot/Restore
+type: bug
+issues: []

server/src/main/java/org/elasticsearch/action/ActionModule.java

@@ -704,7 +704,7 @@ public void initRestHandlers(Supplier<DiscoveryNodes> nodesInCluster) {
         registerHandler.accept(new RestDeleteRepositoryAction());
         registerHandler.accept(new RestVerifyRepositoryAction());
         registerHandler.accept(new RestCleanupRepositoryAction());
-        registerHandler.accept(new RestGetSnapshotsAction());
+        registerHandler.accept(new RestGetSnapshotsAction(threadPool));
         registerHandler.accept(new RestCreateSnapshotAction());
         registerHandler.accept(new RestCloneSnapshotAction());
         registerHandler.accept(new RestRestoreSnapshotAction());

server/src/main/java/org/elasticsearch/action/admin/cluster/snapshots/get/TransportGetSnapshotsAction.java

@@ -86,7 +86,11 @@ public TransportGetSnapshotsAction(
             GetSnapshotsRequest::new,
             indexNameExpressionResolver,
             GetSnapshotsResponse::new,
-            ThreadPool.Names.SAME
+            ThreadPool.Names.MANAGEMENT // Execute this on the management pool because creating the response can become fairly expensive
+                                        // for large repositories in the verbose=false case when there are a lot of indices per snapshot.
+                                        // This is intentionally not using the snapshot_meta pool because that pool is sized rather large
+                                        // to accommodate concurrent IO and could consume excessive CPU resources through concurrent
+                                        // verbose=false requests that are CPU bound only.
         );
         this.repositoriesService = repositoriesService;
     }

server/src/main/java/org/elasticsearch/rest/action/admin/cluster/RestGetSnapshotsAction.java

@@ -13,9 +13,10 @@
 import org.elasticsearch.common.Strings;
 import org.elasticsearch.rest.BaseRestHandler;
 import org.elasticsearch.rest.RestRequest;
+import org.elasticsearch.rest.action.DispatchingRestToXContentListener;
 import org.elasticsearch.rest.action.RestCancellableNodeClient;
-import org.elasticsearch.rest.action.RestToXContentListener;
 import org.elasticsearch.search.sort.SortOrder;
+import org.elasticsearch.threadpool.ThreadPool;
 
 import java.io.IOException;
 import java.util.List;
@@ -31,6 +32,12 @@
  */
 public class RestGetSnapshotsAction extends BaseRestHandler {
 
+    private final ThreadPool threadPool;
+
+    public RestGetSnapshotsAction(ThreadPool threadPool) {
+        this.threadPool = threadPool;
+    }
+
     @Override
     public List<Route> routes() {
         return List.of(new Route(GET, "/_snapshot/{repository}/{snapshot}"));
@@ -76,6 +83,9 @@ public RestChannelConsumer prepareRequest(final RestRequest request, final NodeC
         getSnapshotsRequest.masterNodeTimeout(request.paramAsTime("master_timeout", getSnapshotsRequest.masterNodeTimeout()));
         return channel -> new RestCancellableNodeClient(client, request.getHttpChannel()).admin()
             .cluster()
-            .getSnapshots(getSnapshotsRequest, new RestToXContentListener<>(channel));
+            .getSnapshots(
+                getSnapshotsRequest,
+                new DispatchingRestToXContentListener<>(threadPool.executor(ThreadPool.Names.MANAGEMENT), channel, request)
+            );
     }
 }