We are using RabbitMq with default spring boot configurations. We have a use case in which we want no parallelism for one of the listeners. That is, we want only one thread of the consumer to be running at any given point in time. We want this, because the nature of the use case is such that we want the messages to be consumed in order, thus if there are multiple threads per consumer there can be chances that the messages are processed out of order.
Since, we are using the defaults and have not explicitly tweaked the container, we are using the SimpleMessageListenerContainer. By looking at the documentation I tried fixing the number of consumers using concurrency = "1" . The annotation on the target method looks like this #RabbitListener(queues = ["queue-name"], concurrency = "1").
As per the documentation this should have ensured that there is only consumer thread.
{#link org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer
* SimpleMessageListenerContainer} if this value is a simple integer, it sets a fixed
* number of consumers in the {#code concurrentConsumers} property
2021-10-29 06:11:26.361 INFO 29752 --- [ntContainer#4-1] c.t.t.i.p.s.xxx : Created xxx
2021-10-29 06:11:26.383 INFO 29752 --- [ntContainer#0-1] c.t.t.i.p.s.xxx : Created xxx
ThreadIds to be noted here are [ntContainer#4-1] and [ntContainer#0-1].
So the question is- how can we ensure that there is only one thread per consumer at any given point in time ?
Edit: Adding the code of the consumer class for more context
#ConditionalOnProperty(value = ["rabbitmq.sharebooking.enabled"], havingValue = "true", matchIfMissing = false)
class ShareBookingConsumer #Autowired constructor(
private val shareBookingRepository: ShareBookingRepository,
private val objectMapper: ObjectMapper,
private val shareDtoToShareBookingConverter: ShareBookingDtoToShareBookingConverter
) {
private val logger = LoggerFactory.getLogger(javaClass)
init {
logger.info("start sharebooking created consumer")
}
#RabbitListener(queues = ["tax_engine.share_booking"], concurrency = "1-1", exclusive = true)
#Timed
#Transactional
fun consumeShareBookingCreatedEvent(message: Message) {
try {
consumeShareBookingCreatedEvent(message.body)
} catch (e: Exception) {
throw AmqpRejectAndDontRequeueException(e)
}
}
private fun consumeShareBookingCreatedEvent(event: ByteArray) {
toShareBookingCreationMessageEvent(event).let { creationEvent ->
RmqMetrics.measureEventMetrics(creationEvent)
val shareBooking = shareDtoToShareBookingConverter.convert(creationEvent.data)
val persisted = shareBookingRepository.save(shareBooking)
logger.info("Created shareBooking ${creationEvent.data.id}")
}
}
private fun toShareBookingCreationMessageEvent(event: ByteArray) =
objectMapper.readValue(event, shareBookingCreateEventType)
companion object {
private val shareBookingCreateEventType =
object : TypeReference<RMQMessageEnvelope<ShareBookingCreationDto>>() {}
}
}
Edit: Adding application thread analysis using visualvm
5 threads get created for 5 listeners.
[1]: https://i.stack.imgur.com/gQINE.png
Set concurrency = "1-1". Note that the concurrency of the Listener depends not only on concurrentConsumers, but also on maxConcurrentConsumers:
If you are using a custom factory:
#Bean
public SimpleRabbitListenerContainerFactory rabbitListenerContainerFactory(CachingConnectionFactory cachingConnectionFactory) {
SimpleRabbitListenerContainerFactory factory = new SimpleRabbitListenerContainerFactory();
factory.setConnectionFactory(cachingConnectionFactory);
factory.setConcurrentConsumers(1);
factory.setMaxConcurrentConsumers(1);
return factory;
}
See: https://docs.spring.io/spring-amqp/docs/current/reference/html/#simplemessagelistenercontainer for detail.
EDIT:
I did a simple test, 2 consumers&2 threads:
#RabbitListener(queues = "myQueue111", concurrency = "1-1")
public void handleMessage(Object message) throws InterruptedException {
LOGGER.info("Received message : {} in {}", message, Thread.currentThread().getName());
}
#RabbitListener(queues = "myQueue222", concurrency = "1-1")
public void handleMessag1e(Object message) throws InterruptedException {
LOGGER.info("Received message222 : {} in {}", message, Thread.currentThread().getName());
}
Try this:
#RabbitListener(queues = ["queue-name"], exclusive = true)
See https://docs.spring.io/spring-amqp/docs/current/reference/html/#exclusive-consumer
Related
Spring boot version : 2.7.6
Spring kafka version : 2.8.11
Issue:
I was trying to handle the deserialization issues in code. To handle such issues in code, I created my own class by extending
DefaultErrorHandler
and overriding the public void handleOtherException(Exception thrownException, Consumer<?, ?> consumer, MessageListenerContainer container, boolean batchListener) {}
Sample code as below
public class CustomDefaultErrorHandler extends DefaultErrorHandler {
private static Logger log = LoggerFactory.getLogger(CustomDefaultErrorHandler.class);
#Override
public void handleOtherException(Exception thrownException, Consumer<?, ?> consumer, MessageListenerContainer container, boolean batchListener) {
manageException(thrownException, consumer);
}
private void manageException(Exception ex, Consumer<?, ?> consumer) {
log.error("Error polling message: " + ex.getMessage());
if (ex instanceof RecordDeserializationException) {
RecordDeserializationException rde = (RecordDeserializationException) ex;
consumer.seek(rde.topicPartition(), rde.offset() + 1L);
consumer.commitSync();
} else {
log.error("Exception not handled");
}
}
}
If I use the #RetryableTopic along with #KafkaListener
#RetryableTopic(listenerContainerFactory = "kafkaListenerContainerFactory", backoff = #Backoff(delay = 8000, multiplier = 2.0),
dltStrategy = DltStrategy.FAIL_ON_ERROR
, traversingCauses = "true", autoCreateTopics = "true", numPartitions = "3", replicationFactor = "3",
fixedDelayTopicStrategy = FixedDelayStrategy.MULTIPLE_TOPICS, include = {RetriableException.class, RecoverableDataAccessException.class,
SQLTransientException.class, CallNotPermittedException.class}
)
#KafkaListener(topics = "${topic.name}", groupId = "order", containerFactory = "kafkaListenerContainerFactory", id = "OTR")
public void consumeOTRMessages(ConsumerRecord<String, PayloadsVO> payload, #Header(KafkaHeaders.RECEIVED_TOPIC) String topicName) throws JsonProcessingException {
logger.info("Payload :{}", payload.value());
payloadsService.savePayload(payload.value(), pegasusTopicName);
}
What I saw in debugging the code, #RetryableTopic has its own DefaultErrorHandler configurations in
ListenerContainerFactoryConfigurer
and it stops my custom handler and deserialization process wont stop on issue.
Can you please suggest any way since I wanted to use annotations for retry process in my code
I tried to configured my own implementation of
DefaultErrorHandler
by extending it and configured in
ConcurrentKafkaListenerContainerFactory
It's quite involved, but you should be able to override the RetryTopicComponentFactory bean and override listenerContainerFactoryConfigurer() to return your custom error handler.
That said, deserialization exceptions will go straight to the DLT anyway.
BTW, calling commitSync() here is worthless because there were no records returned by the poll().
I am investigating a bug report in our application (spring boot) regarding the kafka metric kafka.consumer.fetch.manager.records.consumed.total being missing.
The application has two kafka consumers, lets call them query-routing and query-tracking consumers, and they are configured via #KafkaListener annotation and each kafka consumer has it's own instance of ConcurrentKafkaListenerContainerFactory.
The query-router consumer is configured as
#Configuration
#EnableKafka
public class QueryRoutingConfiguration {
#Bean(name = "queryRoutingContainerFactory")
public ConcurrentKafkaListenerContainerFactory<String, RoutingInfo> kafkaListenerContainerFactory(MeterRegistry meterRegistry) {
Map<String, Object> consumerConfigs = new HashMap<>();
// For brevity I removed the configs as they are trivial configs like bootstrap servers and serializers
DefaultKafkaConsumerFactory<String, RoutingInfo> consumerFactory =
new DefaultKafkaConsumerFactory<>(consumerConfigs);
consumerFactory.addListener(new MicrometerConsumerListener<>(meterRegistry));
ConcurrentKafkaListenerContainerFactory<String, RoutingInfo> factory =
new ConcurrentKafkaListenerContainerFactory<>();
factory.setConsumerFactory(consumerFactory);
factory.getContainerProperties().setIdleEventInterval(5000L);
return factory;
}
}
And the query-tracking consumer is configured as:
#Configuration
#EnableKafka
public class QueryTrackingConfiguration {
private static final FixedBackOff NO_ATTEMPTS = new FixedBackOff(Duration.ofSeconds(0).toMillis(), 0L);
#Bean(name = "queryTrackingContainerFactory")
public ConcurrentKafkaListenerContainerFactory<String, QueryTrackingMessage> kafkaListenerContainerFactory(MeterRegistry meterRegistry) {
Map<String, Object> consumerConfigs = new HashMap<>();
// For brevity I removed the configs as they are trivial configs like bootstrap servers and serializers
DefaultKafkaConsumerFactory<String, QueryTrackingMessage> consumerFactory =
new DefaultKafkaConsumerFactory<>(consumerConfigs);
consumerFactory.addListener(new MicrometerConsumerListener<>(meterRegistry));
ConcurrentKafkaListenerContainerFactory<String, QueryTrackingMessage> factory =
new ConcurrentKafkaListenerContainerFactory<>();
factory.setConsumerFactory(consumerFactory);
factory.getContainerProperties().setAckMode(ContainerProperties.AckMode.MANUAL);
factory.setBatchListener(true);
DefaultErrorHandler deusErrorHandler = new DefaultErrorHandler(NO_ATTEMPTS);
factory.setCommonErrorHandler(deusErrorHandler);
return factory;
}
}
The MeterRegistryConfigurator bean configuaration is set as:
#Configuration
public class MeterRegistryConfigurator {
private static final Logger LOG = LoggerFactory.getLogger(MeterRegistryConfigurator.class);
private static final String PREFIX = "dps";
#Bean
MeterRegistryCustomizer<MeterRegistry> meterRegistryCustomizer() {
return registry -> registry.config()
.onMeterAdded(meter -> LOG.info("onMeterAdded: {}", meter.getId().getName()))
.onMeterRemoved(meter -> LOG.info("onMeterRemoved: {}", meter.getId().getName()))
.onMeterRegistrationFailed(
(id, s) -> LOG.info("onMeterRegistrationFailed - id '{}' value '{}'", id.getName(), s))
.meterFilter(PrefixMetricFilter.withPrefix(PREFIX))
.meterFilter(
MeterFilter.deny(id ->
id.getName().startsWith(PREFIX + ".jvm")
|| id.getName().startsWith(PREFIX + ".system")
|| id.getName().startsWith(PREFIX + ".process")
|| id.getName().startsWith(PREFIX + ".logback")
|| id.getName().startsWith(PREFIX + ".tomcat"))
)
.meterFilter(MeterFilter.ignoreTags("host", "host.name"))
.namingConvention(NamingConvention.snakeCase);
}
}
The #KafkaListener for each consumer is set as
#KafkaListener(
id = "query-routing",
idIsGroup = true,
topics = "${query-routing.consumer.topic}",
groupId = "${query-routing.consumer.groupId}",
containerFactory = "queryRoutingContainerFactory")
public void listenForMessages(ConsumerRecord<String, RoutingInfo> record) {
// Handle each record ...
}
and
#KafkaListener(
id = "query-tracking",
idIsGroup = true,
topics = "${query-tracking.consumer.topic}",
groupId = "${query-tracking.consumer.groupId}",
containerFactory = "queryTrackingContainerFactory"
)
public void listenForMessages(List<ConsumerRecord<String, QueryTrackingMessage>> consumerRecords, Acknowledgment ack) {
// Handle each record ...
}
When the application starts up, going to the actuator/prometheus endpoing I can see the metric for both consumers:
# HELP dps_kafka_consumer_fetch_manager_records_consumed_total The total number of records consumed
# TYPE dps_kafka_consumer_fetch_manager_records_consumed_total counter
dps_kafka_consumer_fetch_manager_records_consumed_total{client_id="consumer-qf-query-tracking-consumer-1",kafka_version="3.1.2",spring_id="not.managed.by.Spring.consumer-qf-query-tracking-consumer-1",} 7.0
dps_kafka_consumer_fetch_manager_records_consumed_total{client_id="consumer-QF-Routing-f5d0d9f1-e261-407b-954d-5d217211dee0-2",kafka_version="3.1.2",spring_id="not.managed.by.Spring.consumer-QF-Routing-f5d0d9f1-e261-407b-954d-5d217211dee0-2",} 0.0
But a few seconds later there is a new call to io.micrometer.core.instrument.binder.kafka.KafkaMetrics#checkAndBindMetrics which will remove a set of metrics (including kafka.consumer.fetch.manager.records.consumed.total)
onMeterRegistrationFailed - dps.kafka.consumer.fetch.manager.records.consumed.total string Prometheus requires that all meters with the same name have the same set of tag keys. There is already an existing meter named 'dps.kafka.consumer.fetch.manager.records.consumed.total' containing tag keys [client_id, kafka_version, spring_id]. The meter you are attempting to register has keys [client_id, kafka_version, spring_id, topic].
Going again to actuator/prometheus will only show the metric for the query-routing consumer:
# HELP deus_dps_persistence_kafka_consumer_fetch_manager_records_consumed_total The total number of records consumed for a topic
# TYPE deus_dps_persistence_kafka_consumer_fetch_manager_records_consumed_total counter
deus_dps_persistence_kafka_consumer_fetch_manager_records_consumed_total{client_id="consumer-QF-Routing-0a739a21-4764-411a-9cc6-0e60293b40b4-2",kafka_version="3.1.2",spring_id="not.managed.by.Spring.consumer-QF-Routing-0a739a21-4764-411a-9cc6-0e60293b40b4-2",theKey="routing",topic="QF_query_routing_v1",} 0.0
As you can see above the metric for the query-tracking consumer is gone.
As the log says, The meter you are attempting to register has keys [client_id, kafka_version, spring_id, topic]. The issue is I cannot find where is this metric with a topic key being registered which will trigger io.micrometer.core.instrument.binder.kafka.KafkaMetrics#checkAndBindMetrics which will remove the metric for the query-tracking consumer.
I am using
micrometer-registry-prometheus version 1.9.5
spring boot version 2.7.5
spring kafka (org.springframework.kafka:spring-kafka)
My question is, why does the metric kafka.consumer.fetch.manager.records.consumed.total fails causing it to be removed for the query-tracking consumer and how can I fix it?
I believe this is internal in Micrometer KafkaMetrics.
Periodically, it checks for new metrics; presumably, the topic one shows up after the consumer subscribes to the topic.
#Override
public void bindTo(MeterRegistry registry) {
this.registry = registry;
commonTags = getCommonTags(registry);
prepareToBindMetrics(registry);
checkAndBindMetrics(registry);
VVVVVVVVVVVVVVVVVVVVVVVVVVVVVV
scheduler.scheduleAtFixedRate(() -> checkAndBindMetrics(registry), getRefreshIntervalInMillis(),
getRefreshIntervalInMillis(), TimeUnit.MILLISECONDS);
}
You should be able to write a filter to exclude the one with fewer tags.
I am working on project where I need to validate consumer group is created on topic or not. Is there any way in boldSpring Kafkabold to validate it
Currently, I haven't seen describeConsumerGroups supported in Spring-Kafka KafkaAdmin. So, you may need to create a Kafka AdminClient and call the method by yourself.
E.g: Here, I took advantage of the auto-configuration property class KafkaProperties and autowired it to the service.
#Service
public class KafkaBrokerService implements BrokerService {
private Map<String, Object> configs;
public KafkaBrokerService(KafkaProperties kafkaProperties) { // Autowired
this.configs = kafkaProperties.buildAdminProperties();
}
private AdminClient createAdmin() {
Map<String, Object> configs2 = new HashMap<>(this.configs);
return AdminClient.create(configs2);
}
public SomeDto consumerGroupDescription(String groupId) {
try (AdminClient adminClient = createAdmin()) {
// ConsumerGroup's members
ConsumerGroupDescription consumerGroupDescription = adminClient.describeConsumerGroups(Collections.singletonList(groupId))
.describedGroups().get(groupId).get();
// ConsumerGroup's partitions and the committed offset in each partition
Map<TopicPartition, OffsetAndMetadata> offsets = adminClient.listConsumerGroupOffsets(groupId).partitionsToOffsetAndMetadata().get();
// When you get the information, you can validate it here.
...
} catch (ExecutionException | InterruptedException e) {
//
}
}
}
As described in the documentation, an offset should only be committed when I actually commit (When AckMode.MANUAL_IMMEDIATE or AckMode.MANUAL) or at the end of the listener execution when AckMode.RECORD, however, in the middle of processing the method annotated with #KafkaListener the application goes down, the message is not re-delivered, the application starts reading from the next valid message and this current message is lost (message that was being processed when the application was restarted), how do I achieve the goal of the application reprocessing an uncommitted message when the application is restarted in the middle of processing? I have also tried configuring AUTO_OFFSET_RESET_CONFIG as earliest, latest and none without success in the 3 models. For testing purposes, I created a topic with just one partition, I forced the listener to use the container factory that I define manually.
The springboot-version 2.2.6
#Configuration
class KafkaTestConfiguration {
#Bean
fun producerFactory(): ProducerFactory<String, String> {
return DefaultKafkaProducerFactory(producerConfigs())
}
#Bean
fun consumerFactory(): ConsumerFactory<Any, Any> {
return DefaultKafkaConsumerFactory(consumerConfigs())
}
#Bean
fun producerConfigs(): Map<String, Any> {
val props: MutableMap<String, Any> = HashMap()
props[ProducerConfig.BOOTSTRAP_SERVERS_CONFIG] = "localhost:9094"
props[ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG] = StringSerializer::class.java
props[ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG] = StringSerializer::class.java
return props
}
#Bean
fun consumerConfigs(): Map<String, Any> {
val props: MutableMap<String, Any> = HashMap()
props[ConsumerConfig.BOOTSTRAP_SERVERS_CONFIG] = "localhost:9094"
props[ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG] = StringDeserializer::class.java
props[ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG] = StringDeserializer::class.java
props[ConsumerConfig.MAX_POLL_INTERVAL_MS_CONFIG] = 20000
props[ConsumerConfig.GROUP_ID_CONFIG] = "kafka-retry"
props[ConsumerConfig.AUTO_OFFSET_RESET_CONFIG] = "earliest"
props[ConsumerConfig.ENABLE_AUTO_COMMIT_CONFIG] = "false"
return props
}
#Bean
fun kafkaTemplate(): KafkaTemplate<String, String> {
return KafkaTemplate(producerFactory())
}
#Bean
fun kafkaListenerContainerFactory(): ConcurrentKafkaListenerContainerFactory<Any, Any> {
val factory: ConcurrentKafkaListenerContainerFactory<Any, Any> = ConcurrentKafkaListenerContainerFactory()
factory.consumerFactory = consumerFactory()
factory.consumerFactory.createConsumer()
val containerProperties = factory.containerProperties
containerProperties.isAckOnError = false
containerProperties.ackMode = AckMode.MANUAL_IMMEDIATE
containerProperties.commitLogLevel = LogIfLevelEnabled.Level.INFO
containerProperties.isLogContainerConfig = true
return factory
}
#Component
class KafkaListenerAck {
#KafkaListener(id = "listMsgAckConsumer", topics = ["kafkaListenerTest1"],
groupId = "kafka-retry",
concurrency = "1",
containerFactory = "kafkaListenerContainerFactory"
)
fun onMessage(data: ConsumerRecord<String, String>, acknowledgment: Acknowledgment?) {
println("listMsgAckConsumer1 - topic ${data.topic()} offset ${data.offset()} partition ${data.partition()} message ${data.value()}")
println("If stop container here, the next pool will not deliver the current unconfirmed message")
acknowledgment?.acknowledge()
}
}
The offset will not be committed until the acknowledgment.acknowledge() is called. Set the commitLogLevel container property to DEBUG to see commit activity.
auto.offset.reset only applies if the consumer has never committed an offset (new consumer groups only).
If you can't figure it out from the log; edit the question with the log snippet.
We have a Spring Java application using RabbitMQ, and here is the scenario:
There is a consumer receiving messages from a queue and sending them to another one. We are using "SimpleRabbitListenerContainerFactory" as the container factory, but when sending the messages to the other queue inside a "parallelStream" we've got an IllegalStateException "Cannot determine target ConnectionFactory for lookup key" Exception
When we remove the "parallelStream" it works flawlessly.
public void sendMessage(final StagingMessage stagingMessage, final Long timestamp, final String country) {
final List<TransformedMessage> messages = processMessageList(stagingMessage);
messages.parallelStream().forEach(message -> {
final TransformedMessage transformedMessage = buildMessage(timestamp, ApiConstants.POST_METHOD, country);
myMessageSender.sendQueue(country, transformedMessage);
});
}
Connectio Facotory, where the lookup key is set:
#Configuration
#EnableRabbit
public class RabbitBaseConfig {
#Autowired
private QueueProperties queueProperties;
#Bean
#Primary
public ConnectionFactory connectionFactory(final ConnectionFactory connectionFactoryA, final ConnectionFactory connectionFactoryB) {
final SimpleRoutingConnectionFactory simpleRoutingConnectionFactory = new SimpleRoutingConnectionFactory();
final Map<Object, ConnectionFactory> map = new HashMap<>();
for (final String queue : queueProperties.getAQueueMap().values()) {
map.put("[" + queue + "]", connectionFactoryA);
}
for (final String queue : queueProperties.getBQueueMap().values()) {
map.put("[" + queue + "]", connectionFactoryB);
}
simpleRoutingConnectionFactory.setTargetConnectionFactories(map);
return simpleRoutingConnectionFactory;
}
#Bean
public Jackson2JsonMessageConverter jackson2JsonMessageConverter() {
return new Jackson2JsonMessageConverter();
}
}
Welcome to stack overflow!
You should always show the pertinent code and configuration beans when asking questions like this.
I assume you are using the RoutingConnectionFactory.
It uses a ThreadLocal to store the lookup key so the send has to happen on the same thread that set the key.
You generally should never go asynchronous in a listener anyway; you risk message loss. To increase concurrency, use the concurrency properties on the container.
EDIT
One technique would be to convey the lookup key in a message header:
#Bean
public RabbitTemplate template(ConnectionFactory rcf) {
RabbitTemplate rabbitTemplate = new RabbitTemplate(rcf);
Expression expression = new SpelExpressionParser().parseExpression("messageProperties.headers['cfSelector']");
rabbitTemplate.setSendConnectionFactorySelectorExpression(expression);
return rabbitTemplate;
}
#RabbitListener(queues = "foo")
public void listen1(String in) {
IntStream.range(0, 10)
.parallel()
.mapToObj(i -> in + i)
.forEach(val -> {
this.template.convertAndSend("bar", val.toUpperCase(), msg -> {
msg.getMessageProperties().setHeader("cfSelector", "[bar]");
return msg;
});
});
}