Merge pull request #5909 from sfc-gh-jfu/jfu-cc-request-dbinfo

Change dbinfo broadcast to be explicitly requested by the worker registration message

fdbserver/ClusterController.actor.cpp

@@ -3876,8 +3876,6 @@ ACTOR Future<Void> workerAvailabilityWatch(WorkerInterface worker,
 	        : waitFailureClient(worker.waitFailure, SERVER_KNOBS->WORKER_FAILURE_TIME);
 	cluster->updateWorkerList.set(worker.locality.processId(),
 	                              ProcessData(worker.locality, startingClass, worker.stableAddress()));
-	cluster->updateDBInfoEndpoints.insert(worker.updateServerDBInfo.getEndpoint());
-	cluster->updateDBInfo.trigger();
 	// This switching avoids a race where the worker can be added to id_worker map after the workerAvailabilityWatch
 	// fails for the worker.
 	wait(delay(0));
@@ -4301,6 +4299,8 @@ void registerWorker(RegisterWorkerRequest req, ClusterControllerData* self, Conf
 			    self->id_worker[w.locality.processId()].watcher,
 			    self->id_worker[w.locality.processId()].details.interf.configBroadcastInterface));
 		}
+		self->updateDBInfoEndpoints.insert(w.updateServerDBInfo.getEndpoint());
+		self->updateDBInfo.trigger();
 		checkOutstandingRequests(self);
 	} else if (info->second.details.interf.id() != w.id() || req.generation >= info->second.gen) {
 		if (!info->second.reply.isSet()) {
@@ -4319,6 +4319,10 @@ void registerWorker(RegisterWorkerRequest req, ClusterControllerData* self, Conf
 			info->second.details.interf = w;
 			info->second.watcher = workerAvailabilityWatch(w, newProcessClass, self);
 		}
+		if (req.requestDbInfo) {
+			self->updateDBInfoEndpoints.insert(w.updateServerDBInfo.getEndpoint());
+			self->updateDBInfo.trigger();
+		}
 		if (configBroadcaster != nullptr) {
 			self->addActor.send(
 			    configBroadcaster->registerWorker(req.lastSeenKnobVersion,

fdbserver/WorkerInterface.actor.h

@@ -415,6 +415,7 @@ struct RegisterWorkerRequest {
 	bool degraded;
 	Version lastSeenKnobVersion;
 	ConfigClassSet knobConfigClassSet;
+	bool requestDbInfo;
 
 	RegisterWorkerRequest()
 	  : priorityInfo(ProcessClass::UnsetFit, false, ClusterControllerPriorityInfo::FitnessUnknown), degraded(false) {}
@@ -431,7 +432,8 @@ struct RegisterWorkerRequest {
 	                      ConfigClassSet knobConfigClassSet)
 	  : wi(wi), initialClass(initialClass), processClass(processClass), priorityInfo(priorityInfo),
 	    generation(generation), distributorInterf(ddInterf), ratekeeperInterf(rkInterf), blobManagerInterf(bmInterf),
-	    degraded(degraded), lastSeenKnobVersion(lastSeenKnobVersion), knobConfigClassSet(knobConfigClassSet) {}
+	    degraded(degraded), lastSeenKnobVersion(lastSeenKnobVersion), knobConfigClassSet(knobConfigClassSet),
+	    requestDbInfo(false) {}
 
 	template <class Ar>
 	void serialize(Ar& ar) {
@@ -449,7 +451,8 @@ struct RegisterWorkerRequest {
 		           reply,
 		           degraded,
 		           lastSeenKnobVersion,
-		           knobConfigClassSet);
+		           knobConfigClassSet,
+		           requestDbInfo);
 	}
 };
 

fdbserver/worker.actor.cpp

@@ -522,7 +522,8 @@ ACTOR Future<Void> registrationClient(Reference<AsyncVar<Optional<ClusterControl
                                       Reference<AsyncVar<bool> const> degraded,
                                       Reference<IClusterConnectionRecord> connRecord,
                                       Reference<AsyncVar<std::set<std::string>> const> issues,
-                                      Reference<LocalConfiguration> localConfig) {
+                                      Reference<LocalConfiguration> localConfig,
+                                      Reference<AsyncVar<ServerDBInfo>> dbInfo) {
 	// Keeps the cluster controller (as it may be re-elected) informed that this worker exists
 	// The cluster controller uses waitFailureClient to find out if we die, and returns from registrationReply
 	// (requiring us to re-register) The registration request piggybacks optional distributor interface if it exists.
@@ -533,6 +534,7 @@ ACTOR Future<Void> registrationClient(Reference<AsyncVar<Optional<ClusterControl
 	state Future<Void> cacheProcessFuture;
 	state Future<Void> cacheErrorsFuture;
 	state Optional<double> incorrectTime;
+	state bool firstReg = true;
 	loop {
 		state ClusterConnectionString storedConnectionString;
 		state bool upToDate = true;
@@ -585,6 +587,13 @@ ACTOR Future<Void> registrationClient(Reference<AsyncVar<Optional<ClusterControl
 		}
 
 		state bool ccInterfacePresent = ccInterface->get().present();
+		if (ccInterfacePresent) {
+			request.requestDbInfo = (ccInterface->get().get().id() != dbInfo->get().clusterInterface.id());
+			if (firstReg) {
+				request.requestDbInfo = true;
+				firstReg = false;
+			}
+		}
 		state Future<RegisterWorkerReply> registrationReply =
 		    ccInterfacePresent ? brokenPromiseToNever(ccInterface->get().get().registerWorker.getReply(request))
 		                       : Never();
@@ -1613,7 +1622,8 @@ ACTOR Future<Void> workerServer(Reference<IClusterConnectionRecord> connRecord,
 		                                       degraded,
 		                                       connRecord,
 		                                       issues,
-		                                       localConfig));
+		                                       localConfig,
+		                                       dbInfo));
 
 		if (configDBType != ConfigDBType::DISABLED) {
 			errorForwarders.add(localConfig->consume(interf.configBroadcastInterface));