I have a Reactive Spring Boot application consuming messages from RabbitMQ and persisting them in a (MongoDB) repository:
#RabbitListener(...)
public void processMessage(Message message) {
repository.persist(message).subscribe();
}
Assuming multiple messages arriving in a short period of time, this code could deplete a configured ConnectionPool to the database. If I would receive the messages in a Flux, I could concatMap() them into the db or insert them in buckets of n documents.
That's why I tried to implement a bridge of the given RabbitMQ listener to a self-managed Flux:
#Component
public class QueueListenerController {
private final MyMongoRepository repository;
private final FluxProcessor<Message, Message> fluxProcessor;
private final FluxSink<Message> fluxSink;
public QueueListenerController(MyMongoRepository repository) {
this.repository = repository;
this.fluxProcessor = DirectProcessor.<Message>create().serialize();
this.fluxSink = fluxProcessor.sink();
}
#PostConstruct
private void postConstruct() {
fluxProcessor.concatMap(repository::persist)
.subscribe();
}
#RabbitListener(bindings = #QueueBinding(
value = #Queue(value = "my-queue", durable = "true", autoDelete = "false"),
exchange = #Exchange(value = "amq.direct", durable = "true", autoDelete = "false")
))
public void processMessage(Message message) {
fluxSink.next(message);
}
}
This works locally and for a certain period of time but after some time (I expect 12-24 hours) it stops storing messages in the database, so I'm quite sure I'm doing something wrong.
What would be the correct way of transforming incoming RabbitMQ messages into a Flux of messages?
Related
I want to use chronicle queue to store messages using the high level API as mentioned in the answer of this question. But I also want some kind of key for my messages as mentioned here
1.) Firstly , is this the right/efficient way to read/write using high level API? - Code samples below
2.) How do I separate different category of messages? For example "get me all messages for a particular key , the key in code sample below being ric". Maybe use different topics in the same queue? But how would I do that?
Here's my test code to write to the queue:
public void saveHighLevel(MyInterface obj)
{
try (ChronicleQueue queue = ChronicleQueue.singleBuilder(_location).build()) {
ExcerptAppender appender = queue.acquireAppender();
MyInterface trade = appender.methodWriter(MyInterface.class);
// Write
trade.populate(obj);
}
}
And here's one to read:
public void readHighLevel()
{
try(ChronicleQueue queue = ChronicleQueue.singleBuilder(_location).build()) {
ExcerptTailer tailer = queue.createTailer();
MyInterface container = new MyData();
MethodReader reader = tailer.methodReader(container);
while (reader.readOne()) {
System.out.println(container);
}
}
}
MyInterface:
public interface MyInterface
{
public double getPrice();
public int getSize();
public String getRic();
public void populate(MyInterface obj);
}
Implementation of populate:
public void populate(MyInterface obj)
{
this.price = obj.getPrice();
this.ric = obj.getRic();
this.size = obj.getSize();
}
I found the answer for part (2) of my question in the question of this post.
Essentially by doing:
ChronicleQueue queue = ChronicleQueue.singleBuilder("Topic/SubTopic").build();
where Topic can be substituted with the key I'm looking for.
I have published 50K objects to a specific queue. I have one listener which picks each object and process that. But obviously it will take more time to process all 50k objects. So i want to place 3 more listeners which can parallel process those objects. For this purpose am i need to write two more listener classes? with same code? that will be duplicate of code. Is there any approach we can configure number of listeners we want, so that internally it will create instances for same listener to handle the load?Can any one help me the better way to stand 3 more listeners to handle the load to increase processing.
====Rabbit mq configuration file piece of code=============
#Bean
public SubscriberGeneralQueue1 SubscriberGeneralQueue1(){
return new SubscriberGeneralQueue1();
}
#Bean
public SimpleMessageListenerContainer rpcGeneralReplyMessageListenerContainer(ConnectionFactory connectionFactory,MessageListenerAdapter listenerAdapter1 ) {
SimpleMessageListenerContainer simpleMessageListenerContainer = new SimpleMessageListenerContainer(connectionFactory);
simpleMessageListenerContainer.setQueues(replyQueueRPC());
simpleMessageListenerContainer.setTaskExecutor(taskExecutor());
simpleMessageListenerContainer.setMessageListener(listenerAdapter1);
simpleMessageListenerContainer.setMaxConcurrentConsumers(60);
return simpleMessageListenerContainer;
}
#Bean
#Qualifier("listenerAdapter1")
MessageListenerAdapter listenerAdapter1(SubscriberGeneralQueue1 generalReceiver) {
return new MessageListenerAdapter(generalReceiver, "receivegeneralQueueMessage");
}
===Listener code================
#EnableRabbit
public class SubscriberGeneralQueue1 {
/*#Autowired
#Qualifier("asyncGeneralRabbitTemplate")
private AsyncRabbitTemplate asyncGeneralRabbitTemplate;*/
#Autowired
private ExecutorService executorService;
#Autowired
private GeneralProcess generalProcess;
List <RequestPojo> requestPojoGeneral = new ArrayList<RequestPojo>();
#RabbitHandler
#RabbitListener(containerFactory = "simpleMessageListenerContainerFactory", queues ="BulkSolve_GeneralrequestQueue")
public void subscribeToRequestQueue(#Payload RequestPojo sampleRequestMessage, Message message) throws InterruptedException {
long startTime=System.currentTimeMillis();
//requestPojoGeneral.add(sampleRequestMessage);
//System.out.println("List size issssss:" +requestPojoGeneral.size() );
//generalProcess.processRequestObjectslist(requestPojoGeneral);
generalProcess.processRequestObjects(sampleRequestMessage);
System.out.println("message in general listener is:" + sampleRequestMessage.getDistance());
System.out.println("Message payload is:" + sampleRequestMessage);
System.out.println("Message payload1111 is:" + message );
//return requestPojoGeneral;
}
}
===simplemessagelistenercontainerFactory configuration===========
#Bean
public SimpleRabbitListenerContainerFactory simpleMessageListenerContainerFactory(ConnectionFactory connectionFactory,
SimpleRabbitListenerContainerFactoryConfigurer configurer) {
SimpleRabbitListenerContainerFactory factory = new SimpleRabbitListenerContainerFactory();
factory.setTaskExecutor(taskExecutor());
factory.setMaxConcurrentConsumers(60);
configurer.configure(factory, connectionFactory);
return factory;
}
====Suggested changes=====
#RabbitHandler
#Async
#RabbitListener(containerFactory = "simpleMessageListenerContainerFactory", queues ="BulkSolve_GeneralrequestQueue")
public void subscribeToRequestQueue(#Payload RequestPojo sampleRequestMessage, Message message) throws InterruptedException {
long startTime=System.currentTimeMillis();
//requestPojoGeneral.add(sampleRequestMessage);
//System.out.println("List size issssss:" +requestPojoGeneral.size() );
//generalProcess.processRequestObjectslist(requestPojoGeneral);
generalProcess.processRequestObjects(sampleRequestMessage);
System.out.println("message in general listener is:" + sampleRequestMessage.getDistance());
System.out.println("Message payload is:" + sampleRequestMessage);
System.out.println("Message payload1111 is:" + message );
//return requestPojoGeneral;
}
}
configuration:
#Bean
public SimpleRabbitListenerContainerFactory simpleMessageListenerContainerFactory(ConnectionFactory connectionFactory,
SimpleRabbitListenerContainerFactoryConfigurer configurer) {
SimpleRabbitListenerContainerFactory factory = new SimpleRabbitListenerContainerFactory();
factory.setTaskExecutor(taskExecutor());
factory.setMaxConcurrentConsumers(60);
factory.setConsecutiveActiveTrigger(1);
configurer.configure(factory, connectionFactory);
return factory;
}
#Bean
public SimpleMessageListenerContainer rpcGeneralReplyMessageListenerContainer(ConnectionFactory connectionFactory,MessageListenerAdapter listenerAdapter1 ) {
SimpleMessageListenerContainer simpleMessageListenerContainer = new SimpleMessageListenerContainer(connectionFactory);
simpleMessageListenerContainer.setQueues(replyQueueRPC());
simpleMessageListenerContainer.setTaskExecutor(taskExecutor());
simpleMessageListenerContainer.setMessageListener(listenerAdapter1);
simpleMessageListenerContainer.setMaxConcurrentConsumers(100);
simpleMessageListenerContainer.setConsecutiveActiveTrigger(1);
return simpleMessageListenerContainer;
}
That's can be done with the concurrency option of the ListenerContainer:
Threads from the TaskExecutor configured in the SimpleMessageListenerContainer are used to invoke the MessageListener when a new message is delivered by RabbitMQ Client. If not configured, a SimpleAsyncTaskExecutor is used. If a pooled executor is used, ensure the pool size is sufficient to handle the configured concurrency. With the DirectMessageListenerContainer, the MessageListener is invoked directly on a RabbitMQ Client thread. In this case, the taskExecutor is used for the task that monitors the consumers.
Please, start reading from here: https://docs.spring.io/spring-amqp/docs/current/reference/html/_reference.html#receiving-messages
And also see here: https://docs.spring.io/spring-amqp/docs/current/reference/html/_reference.html#containerAttributes
concurrentConsumers (concurrency) - The number of concurrent consumers to initially start for each listener.
UPDATE
Alright! I see what's going on.
We have there a code like this:
boolean receivedOk = receiveAndExecute(this.consumer); // At least one message received
if (SimpleMessageListenerContainer.this.maxConcurrentConsumers != null) {
if (receivedOk) {
if (isActive(this.consumer)) {
consecutiveIdles = 0;
if (consecutiveMessages++ > SimpleMessageListenerContainer.this.consecutiveActiveTrigger) {
considerAddingAConsumer();
consecutiveMessages = 0;
}
}
}
so, we check for possible parallelism only after the first message is processed. So, in your case it is going to happen after 1 minute.
Another flag to considerAddingAConsumer() is about a consecutiveActiveTrigger option with is this by default:
private static final int DEFAULT_CONSECUTIVE_ACTIVE_TRIGGER = 10;
So, in your case to allow to parallel just exactly the next message you should also configure a :
/**
* If {#link #maxConcurrentConsumers} is greater then {#link #concurrentConsumers}, and
* {#link #maxConcurrentConsumers} has not been reached, specifies the number of
* consecutive cycles when a single consumer was active, in order to consider
* starting a new consumer. If the consumer goes idle for one cycle, the counter is reset.
* This is impacted by the {#link #txSize}.
* Default is 10 consecutive messages.
* #param consecutiveActiveTrigger The number of consecutive receives to trigger a new consumer.
* #see #setMaxConcurrentConsumers(int)
* #see #setStartConsumerMinInterval(long)
* #see #setTxSize(int)
*/
public final void setConsecutiveActiveTrigger(int consecutiveActiveTrigger) {
Assert.isTrue(consecutiveActiveTrigger > 0, "'consecutiveActiveTrigger' must be > 0");
this.consecutiveActiveTrigger = consecutiveActiveTrigger;
}
to 1. Because 0 is not going to work anyway.
For better performance you also may consider to make your subscribeToRequestQueue() with the #Async to really hand off the processing from the consumer thread to some other to avoid that 1 minute to wait for one more consumer to start.
I'm working on a project where we are polling files from a sftp server and streaming it out into a object on the rabbitmq queue. Now when the rabbitmq is down it still polls and deletes the file from the server and losses the file while sending it on queue when rabbitmq is down. I'm using ExpressionEvaluatingRequestHandlerAdvice to remove the file on successful transformation. My code looks like this:
#Bean
public SessionFactory<ChannelSftp.LsEntry> sftpSessionFactory() {
DefaultSftpSessionFactory factory = new DefaultSftpSessionFactory(true);
factory.setHost(sftpProperties.getSftpHost());
factory.setPort(sftpProperties.getSftpPort());
factory.setUser(sftpProperties.getSftpPathUser());
factory.setPassword(sftpProperties.getSftpPathPassword());
factory.setAllowUnknownKeys(true);
return new CachingSessionFactory<>(factory);
}
#Bean
public SftpRemoteFileTemplate sftpRemoteFileTemplate() {
return new SftpRemoteFileTemplate(sftpSessionFactory());
}
#Bean
#InboundChannelAdapter(channel = TransformerChannel.TRANSFORMER_OUTPUT, autoStartup = "false",
poller = #Poller(value = "customPoller"))
public MessageSource<InputStream> sftpMessageSource() {
SftpStreamingMessageSource messageSource = new SftpStreamingMessageSource(sftpRemoteFileTemplate,
null);
messageSource.setRemoteDirectory(sftpProperties.getSftpDirPath());
messageSource.setFilter(new SftpPersistentAcceptOnceFileListFilter(new SimpleMetadataStore(),
"streaming"));
messageSource.setFilter(new SftpSimplePatternFileListFilter("*.txt"));
return messageSource;
}
#Bean
#Transformer(inputChannel = TransformerChannel.TRANSFORMER_OUTPUT,
outputChannel = SFTPOutputChannel.SFTP_OUTPUT,
adviceChain = "deleteAdvice")
public org.springframework.integration.transformer.Transformer transformer() {
return new SFTPTransformerService("UTF-8");
}
#Bean
public ExpressionEvaluatingRequestHandlerAdvice deleteAdvice() {
ExpressionEvaluatingRequestHandlerAdvice advice = new ExpressionEvaluatingRequestHandlerAdvice();
advice.setOnSuccessExpressionString(
"#sftpRemoteFileTemplate.remove(headers['file_remoteDirectory'] + headers['file_remoteFile'])");
advice.setPropagateEvaluationFailures(false);
return advice;
}
I don't want the files to get removed/polled from the remote sftp server when the rabbitmq server is down. How can i achieve this ?
UPDATE
Apologies for not mentioning that I'm using spring cloud stream rabbit binder. And here is the transformer service:
public class SFTPTransformerService extends StreamTransformer {
public SFTPTransformerService(String charset) {
super(charset);
}
#Override
protected Object doTransform(Message<?> message) throws Exception {
String fileName = message.getHeaders().get("file_remoteFile", String.class);
Object fileContents = super.doTransform(message);
return new customFileDTO(fileName, (String) fileContents);
}
}
UPDATE-2
I added TransactionSynchronizationFactory on the customPoller as suggested. Now it doesn't poll file when rabbit server is down, but when the server is up, it keeps on polling the same file over and over again!! I cannot figure it out why? I guess i cannot use PollerSpec cause im on 4.3.2 version.
#Bean(name = "customPoller")
public PollerMetadata pollerMetadataDTX(StartStopTrigger startStopTrigger,
CustomTriggerAdvice customTriggerAdvice) {
PollerMetadata pollerMetadata = new PollerMetadata();
pollerMetadata.setAdviceChain(Collections.singletonList(customTriggerAdvice));
pollerMetadata.setTrigger(startStopTrigger);
pollerMetadata.setMaxMessagesPerPoll(Long.valueOf(sftpProperties.getMaxMessagePoll()));
ExpressionEvaluatingTransactionSynchronizationProcessor syncProcessor =
new ExpressionEvaluatingTransactionSynchronizationProcessor();
syncProcessor.setBeanFactory(applicationContext.getAutowireCapableBeanFactory());
syncProcessor.setBeforeCommitChannel(
applicationContext.getBean(TransformerChannel.TRANSFORMER_OUTPUT, MessageChannel.class));
syncProcessor
.setAfterCommitChannel(
applicationContext.getBean(SFTPOutputChannel.SFTP_OUTPUT, MessageChannel.class));
syncProcessor.setAfterCommitExpression(new SpelExpressionParser().parseExpression(
"#sftpRemoteFileTemplate.remove(headers['file_remoteDirectory'] + headers['file_remoteFile'])"));
DefaultTransactionSynchronizationFactory defaultTransactionSynchronizationFactory =
new DefaultTransactionSynchronizationFactory(syncProcessor);
pollerMetadata.setTransactionSynchronizationFactory(defaultTransactionSynchronizationFactory);
return pollerMetadata;
}
I don't know if you need this info but my CustomTriggerAdvice and StartStopTrigger looks like this :
#Component
public class CustomTriggerAdvice extends AbstractMessageSourceAdvice {
#Autowired private StartStopTrigger startStopTrigger;
#Override
public boolean beforeReceive(MessageSource<?> source) {
return true;
}
#Override
public Message<?> afterReceive(Message<?> result, MessageSource<?> source) {
if (result == null) {
if (startStopTrigger.getStart()) {
startStopTrigger.stop();
}
} else {
if (!startStopTrigger.getStart()) {
startStopTrigger.stop();
}
}
return result;
}
}
public class StartStopTrigger implements Trigger {
private PeriodicTrigger startTrigger;
private boolean start;
public StartStopTrigger(PeriodicTrigger startTrigger, boolean start) {
this.startTrigger = startTrigger;
this.start = start;
}
#Override
public Date nextExecutionTime(TriggerContext triggerContext) {
if (!start) {
return null;
}
start = true;
return startTrigger.nextExecutionTime(triggerContext);
}
public void stop() {
start = false;
}
public void start() {
start = true;
}
public boolean getStart() {
return this.start;
}
}
Well, would be great to see what your SFTPTransformerService and determine how it is possible to perform an onSuccessExpression when there should be an exception in case of down broker.
You also should not only throw an exception do not perform delete, but consider to add a RequestHandlerRetryAdvice to re-send the file to the RabbitMQ: https://docs.spring.io/spring-integration/docs/5.0.6.RELEASE/reference/html/messaging-endpoints-chapter.html#retry-advice
UPDATE
So, well, since Gary guessed that you use Spring Cloud Stream to send message to the Rabbit Binder after your internal process (very sad that you didn't share that information originally), you need to take a look to the Binder error handling on the matter: https://docs.spring.io/spring-cloud-stream/docs/Elmhurst.RELEASE/reference/htmlsingle/#_retry_with_the_rabbitmq_binder
And that is true that ExpressionEvaluatingRequestHandlerAdvice is applied only for the SFTPTransformerService and nothing more. The downstream error (in the Binder) is not included in this process already.
UPDATE 2
Yeah... I think Gary is right, and we don't have choice unless configure a TransactionSynchronizationFactory on the customPoller level instead of that ExpressionEvaluatingRequestHandlerAdvice: ExpressionEvaluatingRequestHandlerAdvice .
The DefaultTransactionSynchronizationFactory can be configured with the ExpressionEvaluatingTransactionSynchronizationProcessor, which has similar goal as the mentioned ExpressionEvaluatingRequestHandlerAdvice, but on the transaction level which will include your process starting with the SFTP Channel Adapter and ending on the Rabbit Binder level with the send to AMQP attempts.
See Reference Manual for more information: https://docs.spring.io/spring-integration/reference/html/transactions.html#transaction-synchronization.
The point with the ExpressionEvaluatingRequestHandlerAdvice (and any AbstractRequestHandlerAdvice) that they have a boundary only around handleRequestMessage() method, therefore only during the component they are declared.
Hi I am new to Storm and Kafka.
I am using storm 1.0.1 and kafka 0.10.0
we have a kafkaspout that would receive java bean from kafka topic.
I have spent several hours digging to find the right approach for that.
Found few articles which are useful but none of the approaches worked for me so far.
Following is my codes:
StormTopology:
public class StormTopology {
public static void main(String[] args) throws Exception {
//Topo test /zkroot test
if (args.length == 4) {
System.out.println("started");
BrokerHosts hosts = new ZkHosts("localhost:2181");
SpoutConfig kafkaConf1 = new SpoutConfig(hosts, args[1], args[2],
args[3]);
kafkaConf1.zkRoot = args[2];
kafkaConf1.useStartOffsetTimeIfOffsetOutOfRange = true;
kafkaConf1.startOffsetTime = kafka.api.OffsetRequest.LatestTime();
kafkaConf1.scheme = new SchemeAsMultiScheme(new KryoScheme());
KafkaSpout kafkaSpout1 = new KafkaSpout(kafkaConf1);
System.out.println("started");
ShuffleBolt shuffleBolt = new ShuffleBolt(args[1]);
AnalysisBolt analysisBolt = new AnalysisBolt(args[1]);
TopologyBuilder topologyBuilder = new TopologyBuilder();
topologyBuilder.setSpout("kafkaspout", kafkaSpout1, 1);
//builder.setBolt("counterbolt2", countbolt2, 3).shuffleGrouping("kafkaspout");
//This is for field grouping in bolt we need two bolt for field grouping or it wont work
topologyBuilder.setBolt("shuffleBolt", shuffleBolt, 3).shuffleGrouping("kafkaspout");
topologyBuilder.setBolt("analysisBolt", analysisBolt, 5).fieldsGrouping("shuffleBolt", new Fields("trip"));
Config config = new Config();
config.registerSerialization(VehicleTrip.class, VehicleTripKyroSerializer.class);
config.setDebug(true);
config.setNumWorkers(1);
LocalCluster cluster = new LocalCluster();
cluster.submitTopology(args[0], config, topologyBuilder.createTopology());
// StormSubmitter.submitTopology(args[0], config,
// builder.createTopology());
} else {
System.out
.println("Insufficent Arguements - topologyName kafkaTopic ZKRoot ID");
}
}
}
I am serializing the data at kafka using kryo
KafkaProducer:
public class StreamKafkaProducer {
private static Producer producer;
private final Properties props = new Properties();
private static final StreamKafkaProducer KAFKA_PRODUCER = new StreamKafkaProducer();
private StreamKafkaProducer(){
props.put("bootstrap.servers", "localhost:9092");
props.put("acks", "all");
props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");
props.put("value.serializer", "com.abc.serializer.MySerializer");
producer = new org.apache.kafka.clients.producer.KafkaProducer(props);
}
public static StreamKafkaProducer getStreamKafkaProducer(){
return KAFKA_PRODUCER;
}
public void produce(String topic, VehicleTrip vehicleTrip){
ProducerRecord<String,VehicleTrip> producerRecord = new ProducerRecord<>(topic,vehicleTrip);
producer.send(producerRecord);
//producer.close();
}
public static void closeProducer(){
producer.close();
}
}
Kyro Serializer:
public class DataKyroSerializer extends Serializer<Data> implements Serializable {
#Override
public void write(Kryo kryo, Output output, VehicleTrip vehicleTrip) {
output.writeLong(data.getStartedOn().getTime());
output.writeLong(data.getEndedOn().getTime());
}
#Override
public Data read(Kryo kryo, Input input, Class<VehicleTrip> aClass) {
Data data = new Data();
data.setStartedOn(new Date(input.readLong()));
data.setEndedOn(new Date(input.readLong()));
return data;
}
I need to get the data back to the Data bean.
As per few articles I need to provide with a custom scheme and make it part of topology but till now I have no luck
Code for Bolt and Scheme
Scheme:
public class KryoScheme implements Scheme {
private ThreadLocal<Kryo> kryos = new ThreadLocal<Kryo>() {
protected Kryo initialValue() {
Kryo kryo = new Kryo();
kryo.addDefaultSerializer(Data.class, new DataKyroSerializer());
return kryo;
};
};
#Override
public List<Object> deserialize(ByteBuffer ser) {
return Utils.tuple(kryos.get().readObject(new ByteBufferInput(ser.array()), Data.class));
}
#Override
public Fields getOutputFields( ) {
return new Fields( "data" );
}
}
and bolt:
public class AnalysisBolt implements IBasicBolt {
/**
*
*/
private static final long serialVersionUID = 1L;
private String topicname = null;
public AnalysisBolt(String topicname) {
this.topicname = topicname;
}
public void prepare(Map stormConf, TopologyContext topologyContext) {
System.out.println("prepare");
}
public void execute(Tuple input, BasicOutputCollector collector) {
System.out.println("execute");
Fields fields = input.getFields();
try {
JSONObject eventJson = (JSONObject) JSONSerializer.toJSON((String) input
.getValueByField(fields.get(1)));
String StartTime = (String) eventJson.get("startedOn");
String EndTime = (String) eventJson.get("endedOn");
String Oid = (String) eventJson.get("_id");
int V_id = (Integer) eventJson.get("vehicleId");
//call method getEventForVehicleWithinTime(Long vehicleId, Date startTime, Date endTime)
System.out.println("==========="+Oid+"| "+V_id+"| "+StartTime+"| "+EndTime);
} catch (Exception e) {
e.printStackTrace();
}
}
but if I submit the storm topology i am getting error:
java.lang.IllegalStateException: Spout 'kafkaspout' contains a
non-serializable field of type com.abc.topology.KryoScheme$1, which
was instantiated prior to topology creation.
com.minda.iconnect.topology.KryoScheme$1 should be instantiated within
the prepare method of 'kafkaspout at the earliest.
Appreciate help to debug the issue and guide to right path.
Thanks
Your ThreadLocal is not Serializable. The preferable solution would be to make your serializer both Serializable and threadsafe. If this is not possible, then I see 2 alternatives since there is no prepare method as you would get in a bolt.
Declare it as static, which is inherently transient.
Declare it transient and access it via a private get method. Then you can initialize the variable on first access.
Within the Storm lifecycle, the topology is instantiated and then serialized to byte format to be stored in ZooKeeper, prior to the topology being executed. Within this step, if a spout or bolt within the topology has an initialized unserializable property, serialization will fail.
If there is a need for a field that is unserializable, initialize it within the bolt or spout's prepare method, which is run after the topology is delivered to the worker.
Source: Best Practices for implementing Apache Storm
I am learning about Spring-Integration and have a basic understanding about Gateway and Service-Activators. I love the concept of Gateway. Spring Integration generates the proxy for gateway at run-time. This proxy hides all the messaging details from the consumer of the gateway. In addition, the generated proxy might also be co-relating request and reply.
With the objective of learning, I set out to implement request and reply correlation using raw Spring Integration features and not using Gateway. I am able to set the correlation identifier in the request header, but not able to specify correlation identifier while receiving reply for the channel. The following (at the end of the question) is the code snippet for the same. Also how does the correlation stuff works against a message broker (e.g. RabbitMQ)? Does RabbitMQ provides an ability to retrieve a message with a specific header (correlation identifier) in it?
public class RemoteProxyCalculatorService implements CalculatorService
{
public int Square(int n)
{
UUID uuid = SendRequest(n, "squareRequestChannel");
int squareOfn = ReceiveReply("squareReplyChannel", uuid);
return squareOfn;
}
private <T> UUID SendRequest(T payload, String requestChannel)
{
UUID requestID = UUID.randomUUID();
Message<T> inputMessage = MessageBuilder.withPayload(payload)
.setCorrelationId(requestID)
.build();
MessageChannel channel = (MessageChannel)context.getBean(requestChannel, MessageChannel.class);
channel.send(inputMessage);
return requestID;
}
#SuppressWarnings("unchecked")
private <T> T ReceiveReply(String replyChannel, UUID requestID)
{
//How to consume requestID so as to receive only the reply related to the request posted by this thread
PollableChannel channel = (PollableChannel)context.getBean(replyChannel);
Message<?> groupMessage = channel.receive();
return (T)groupMessage.getPayload();
}
private ClassPathXmlApplicationContext context;
}
Thanks.
The simplest way to correlate within an app doesn't even require a correlationId header. Instead you can create a QueueChannel instance (that you don't share) and provide that as s the replyChannel header on the Message you send. Whatever downstream component ultimately responds, it will find that header in the Message.
Regarding RabbitMQ, our outbound-gateway simply applies a similar technique, but using the replyTo property of the AMQP Message.
Hope that helps.
-Mark
Problem is with common reply channel. The solution (Mark suggested the similar) will look like this.
public class RemoteProxyCalculatorService
{
public int Square(int n)
{
PollableChannel replyChannel = SendRequest(n, "squareRequestChannel");
int squareOfn = ReceiveReply(replyChannel);
return squareOfn;
}
private <T> PollableChannel SendRequest(T payload, String requestChannel)
{
UUID requestID = UUID.randomUUID();
QueueChannel replyQueueChannel = new QueueChannel();
Message<T> inputMessage = MessageBuilder.withPayload(payload)
.setCorrelationId(requestID)
.setReplyChannel(replyQueueChannel)
.build();
MessageChannel channel = context.getBean(requestChannel, MessageChannel.class);
channel.send(inputMessage);
return replyQueueChannel;
}
#SuppressWarnings("unchecked")
private <T> T ReceiveReply(PollableChannel replyChannel)
{
Message<?> groupMessage = replyChannel.receive();
return (T) groupMessage.getPayload();
}
private ClassPathXmlApplicationContext context;
}
If you want to use common reply channel then I think this is what you are looking for.
public class RemoteProxyCalculatorService
{
public int Square(int n)
{
PollableChannel replyChannel = SendRequest(n, "squareRequestChannel");
int squareOfn = ReceiveReply(replyChannel);
return squareOfn;
}
private <T> PollableChannel SendRequest(T payload, String requestChannel)
{
UUID requestID = UUID.randomUUID();
Message<T> inputMessage = MessageBuilder.withPayload(payload)
.setCorrelationId(requestID)
.setReplyChannel(myMessageHandler.getSubscribedChannel())
.build();
// Create a Pollable channel for two things
// 1. Pollable channel is where this thread should look for reply.
QueueChannel replyQueueChannel = new QueueChannel();
// 2. Message Handler will send reply to this Pollable channel once it receives the reply using correlation Id.
myMessageHandler.add(requestID, replyQueueChannel);
MessageChannel channel = context.getBean(requestChannel, MessageChannel.class);
channel.send(inputMessage);
return replyQueueChannel;
}
#SuppressWarnings("unchecked")
private <T> T ReceiveReply(PollableChannel replyChannel)
{
Message<?> groupMessage = replyChannel.receive();
return (T) groupMessage.getPayload();
}
private ClassPathXmlApplicationContext context;
#Autowired
private MyMessageHandler myMessageHandler;
}
/**
* Message Handler
*
*/
public class MyMessageHandler implements MessageHandler
{
private final Map<Object, MessageChannel> idChannelsMap = new TreeMap<>();
private final Object lock = new Object();
private final SubscribableChannel subscribedChannel;
public MyMessageHandler(SubscribableChannel subscribedChannel)
{
this.subscribedChannel = subscribedChannel;
}
#Override
public void handleMessage(Message<?> message) throws MessagingException
{
synchronized (lock)
{
this.idChannelsMap.get(message.getHeaders().getCorrelationId()).send(message);
this.idChannelsMap.remove(message.getHeaders().getCorrelationId());
}
}
public void add(Object correlationId, MessageChannel messageChannel)
{
synchronized (lock)
{
this.idChannelsMap.put(correlationId, messageChannel);
}
}
public SubscribableChannel getSubscribedChannel()
{
return subscribedChannel;
}
}