How do I prevent ListenerExecutionFailedException: Listener threw exception - spring-integration

Question

What do I need to do to prevent the following Exception which is presumably thrown by RabbitMQ.
org.springframework.amqp.rabbit.listener.exception.ListenerExecutionFailedException: Listener threw exception
at org.springframework.amqp.rabbit.listener.AbstractMessageListenerContainer.wrapToListenerExecutionFailedExceptionIfNeeded(AbstractMessageListenerContainer.java:877)
at org.springframework.amqp.rabbit.listener.AbstractMessageListenerContainer.doInvokeListener(AbstractMessageListenerContainer.java:787)
at org.springframework.amqp.rabbit.listener.AbstractMessageListenerContainer.invokeListener(AbstractMessageListenerContainer.java:707)
at org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer.access$001(SimpleMessageListenerContainer.java:98)
at org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer$1.invokeListener(SimpleMessageListenerContainer.java:189)
at org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer.invokeListener(SimpleMessageListenerContainer.java:1236)
at org.springframework.amqp.rabbit.listener.AbstractMessageListenerContainer.executeListener(AbstractMessageListenerContainer.java:688)
at org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer.doReceiveAndExecute(SimpleMessageListenerContainer.java:1190)
at org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer.receiveAndExecute(SimpleMessageListenerContainer.java:1174)
at org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer.access$1200(SimpleMessageListenerContainer.java:98)
at org.springframework.amqp.rabbit.listener.SimpleMessageListenerContainer$AsyncMessageProcessingConsumer.run(SimpleMessageListenerContainer.java:1363)
at java.lang.Thread.run(Thread.java:748)
Caused by: org.springframework.messaging.MessageDeliveryException: failed to send Message to channel 'amqpLaunchSpringBatchJobFlow.channel#0'; nested exception is jp.ixam_drive.batch.service.JobExecutionRuntimeException: Failed to start job with name ads-insights-import and parameters {accessToken=<ACCESS_TOKEN>, id=act_1234567890, classifier=stats, report_run_id=1482330625184792, job_request_id=32}
at org.springframework.integration.channel.AbstractMessageChannel.send(AbstractMessageChannel.java:449)
at org.springframework.integration.channel.AbstractMessageChannel.send(AbstractMessageChannel.java:373)
at org.springframework.messaging.core.GenericMessagingTemplate.doSend(GenericMessagingTemplate.java:115)
at org.springframework.messaging.core.GenericMessagingTemplate.doSend(GenericMessagingTemplate.java:45)
at org.springframework.messaging.core.AbstractMessageSendingTemplate.send(AbstractMessageSendingTemplate.java:105)
at org.springframework.integration.endpoint.MessageProducerSupport.sendMessage(MessageProducerSupport.java:171)
at org.springframework.integration.amqp.inbound.AmqpInboundChannelAdapter.access$400(AmqpInboundChannelAdapter.java:45)
at org.springframework.integration.amqp.inbound.AmqpInboundChannelAdapter$1.onMessage(AmqpInboundChannelAdapter.java:95)
at org.springframework.amqp.rabbit.listener.AbstractMessageListenerContainer.doInvokeListener(AbstractMessageListenerContainer.java:784)
... 10 common frames omitted
Caused by: jp.ixam_drive.batch.service.JobExecutionRuntimeException: Failed to start job with name ads-insights-import and parameters {accessToken=<ACCESS_TOKEN>, id=act_1234567890, classifier=stats, report_run_id=1482330625184792, job_request_id=32}
at jp.ixam_drive.facebook.SpringBatchLauncher.launchJob(SpringBatchLauncher.java:42)
at jp.ixam_drive.facebook.AmqpBatchLaunchIntegrationFlows.lambda$amqpLaunchSpringBatchJobFlow$1(AmqpBatchLaunchIntegrationFlows.java:71)
at org.springframework.integration.dispatcher.AbstractDispatcher.tryOptimizedDispatch(AbstractDispatcher.java:116)
at org.springframework.integration.dispatcher.UnicastingDispatcher.doDispatch(UnicastingDispatcher.java:148)
at org.springframework.integration.dispatcher.UnicastingDispatcher.dispatch(UnicastingDispatcher.java:121)
at org.springframework.integration.channel.AbstractSubscribableChannel.doSend(AbstractSubscribableChannel.java:89)
at org.springframework.integration.channel.AbstractMessageChannel.send(AbstractMessageChannel.java:423)
... 18 common frames omitted
Caused by: org.springframework.batch.core.repository.JobInstanceAlreadyCompleteException: A job instance already exists and is complete for parameters={accessToken=<ACCESS_TOKEN>, id=act_1234567890, classifier=stats, report_run_id=1482330625184792, job_request_id=32}. If you want to run this job again, change the parameters.
at org.springframework.batch.core.repository.support.SimpleJobRepository.createJobExecution(SimpleJobRepository.java:126)
at sun.reflect.GeneratedMethodAccessor193.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at org.springframework.aop.support.AopUtils.invokeJoinpointUsingReflection(AopUtils.java:333)
at org.springframework.aop.framework.ReflectiveMethodInvocation.invokeJoinpoint(ReflectiveMethodInvocation.java:190)
at org.springframework.aop.framework.ReflectiveMethodInvocation.proceed(ReflectiveMethodInvocation.java:157)
at org.springframework.transaction.interceptor.TransactionInterceptor$1.proceedWithInvocation(TransactionInterceptor.java:99)
at org.springframework.transaction.interceptor.TransactionAspectSupport.invokeWithinTransaction(TransactionAspectSupport.java:282)
at org.springframework.transaction.interceptor.TransactionInterceptor.invoke(TransactionInterceptor.java:96)
at org.springframework.aop.framework.ReflectiveMethodInvocation.proceed(ReflectiveMethodInvocation.java:179)
at org.springframework.batch.core.repository.support.AbstractJobRepositoryFactoryBean$1.invoke(AbstractJobRepositoryFactoryBean.java:172)
at org.springframework.aop.framework.ReflectiveMethodInvocation.proceed(ReflectiveMethodInvocation.java:179)
at org.springframework.aop.framework.JdkDynamicAopProxy.invoke(JdkDynamicAopProxy.java:213)
at com.sun.proxy.$Proxy125.createJobExecution(Unknown Source)
at org.springframework.batch.core.launch.support.SimpleJobLauncher.run(SimpleJobLauncher.java:125)
at jp.ixam_drive.batch.service.JobOperationsService.launch(JobOperationsService.java:64)
at jp.ixam_drive.facebook.SpringBatchLauncher.launchJob(SpringBatchLauncher.java:37)
... 24 common frames omitted
when I have 2 instances of Spring Boot application both of which run the following code in parallel to execute Spring Batch Jobs?
#Configuration
#Conditional(AmqpBatchLaunchCondition.class)
#Slf4j
public class AmqpAsyncAdsInsightsConfiguration {
#Autowired
ObjectMapper objectMapper;
#Value("${batch.launch.amqp.routing-keys.async-insights}")
String routingKey;
#Bean
public IntegrationFlow amqpOutboundAsyncAdsInsights(AmqpTemplate amqpTemplate) {
return IntegrationFlows.from("async_ads_insights")
.<JobParameters, byte[]>transform(SerializationUtils::serialize)
.handle(Amqp.outboundAdapter(amqpTemplate).routingKey(routingKey)).get();
}
#Bean
public IntegrationFlow amqpAdsInsightsAsyncJobRequestFlow(FacebookMarketingServiceProvider serviceProvider,
JobParametersToApiParametersTransformer transformer, ConnectionFactory connectionFactory) {
return IntegrationFlows.from(Amqp.inboundAdapter(connectionFactory, routingKey))
.<byte[], JobParameters>transform(SerializationUtils::deserialize)
.<JobParameters, ApiParameters>transform(transformer)
.<ApiParameters>handle((payload, header) -> {
String accessToken = (String) header.get("accessToken");
String id = (String) header.get("object_id");
FacebookMarketingApi api = serviceProvider.getApi(accessToken);
String reportRunId = api.asyncRequestOperations().getReportRunId(id, payload.toMap());
ObjectNode objectNode = objectMapper.createObjectNode();
objectNode.put("accessToken", accessToken);
objectNode.put("id", id);
objectNode.put("report_run_id", reportRunId);
objectNode.put("classifier", (String) header.get("classifier"));
objectNode.put("job_request_id", (Long) header.get("job_request_id"));
return serialize(objectNode);
}).channel("ad_report_run_polling_channel").get();
}
#SneakyThrows
private String serialize(JsonNode jsonNode) {
return objectMapper.writeValueAsString(jsonNode);
}
}
#Configuration
#Conditional(AmqpBatchLaunchCondition.class)
#Slf4j
public class AmqpBatchLaunchIntegrationFlows {
#Autowired
SpringBatchLauncher batchLauncher;
#Value("${batch.launch.amqp.routing-keys.job-launch}")
String routingKey;
#Bean(name = "batch_launch_channel")
public MessageChannel batchLaunchChannel() {
return MessageChannels.executor(Executors.newSingleThreadExecutor()).get();
}
#Bean
public IntegrationFlow amqpOutbound(AmqpTemplate amqpTemplate,
#Qualifier("batch_launch_channel") MessageChannel batchLaunchChannel) {
return IntegrationFlows.from(batchLaunchChannel)
.<JobParameters, byte[]>transform(SerializationUtils::serialize)
.handle(Amqp.outboundAdapter(amqpTemplate).routingKey(routingKey)).get();
}
#Bean
public IntegrationFlow amqpLaunchSpringBatchJobFlow(ConnectionFactory connectionFactory) {
return IntegrationFlows.from(Amqp.inboundAdapter(connectionFactory, routingKey))
.handle(message -> {
String jobName = (String) message.getHeaders().get("job_name");
byte[] bytes = (byte[]) message.getPayload();
JobParameters jobParameters = SerializationUtils.deserialize(bytes);
batchLauncher.launchJob(jobName, jobParameters);
}).get();
}
}
#Configuration
#Slf4j
public class AsyncAdsInsightsConfiguration {
#Value("${batch.core.pool.size}")
public Integer batchCorePoolSize;
#Value("${ixam_drive.facebook.api.ads-insights.async-poll-interval}")
public String asyncPollInterval;
#Autowired
ObjectMapper objectMapper;
#Autowired
private DataSource dataSource;
#Bean(name = "async_ads_insights")
public MessageChannel adsInsightsAsyncJobRequestChannel() {
return MessageChannels.direct().get();
}
#Bean(name = "ad_report_run_polling_channel")
public MessageChannel adReportRunPollingChannel() {
return MessageChannels.executor(Executors.newFixedThreadPool(batchCorePoolSize)).get();
}
#Bean
public IntegrationFlow adReportRunPollingLoopFlow(FacebookMarketingServiceProvider serviceProvider) {
return IntegrationFlows.from(adReportRunPollingChannel())
.<String>handle((payload, header) -> {
ObjectNode jsonNode = deserialize(payload);
String accessToken = jsonNode.get("accessToken").asText();
String reportRunId = jsonNode.get("report_run_id").asText();
try {
AdReportRun adReportRun = serviceProvider.getApi(accessToken)
.fetchObject(reportRunId, AdReportRun.class);
log.debug("ad_report_run: {}", adReportRun);
return jsonNode.set("ad_report_run", objectMapper.valueToTree(adReportRun));
} catch (Exception e) {
log.error("failed while polling for ad_report_run.id: {}", reportRunId);
throw new RuntimeException(e);
}
}).<JsonNode, Boolean>route(payload -> {
JsonNode adReportRun = payload.get("ad_report_run");
return adReportRun.get("async_percent_completion").asInt() == 100 &&
"Job Completed".equals(adReportRun.get("async_status").asText());
}, rs -> rs.subFlowMapping(true,
f -> f.transform(JsonNode.class,
source -> {
JobParametersBuilder jobParametersBuilder = new JobParametersBuilder();
jobParametersBuilder
.addString("accessToken", source.get("accessToken").asText());
jobParametersBuilder.addString("id", source.get("id").asText());
jobParametersBuilder
.addString("classifier", source.get("classifier").asText());
jobParametersBuilder
.addLong("report_run_id", source.get("report_run_id").asLong());
jobParametersBuilder
.addLong("job_request_id", source.get("job_request_id").asLong());
return jobParametersBuilder.toJobParameters();
}).channel("batch_launch_channel"))
.subFlowMapping(false,
f -> f.transform(JsonNode.class, this::serialize)
.<String>delay("delay", asyncPollInterval, c -> c.transactional()
.messageStore(jdbcMessageStore()))
.channel(adReportRunPollingChannel()))).get();
}
#SneakyThrows
private String serialize(JsonNode jsonNode) {
return objectMapper.writeValueAsString(jsonNode);
}
#SneakyThrows
private ObjectNode deserialize(String payload) {
return objectMapper.readerFor(ObjectNode.class).readValue(payload);
}
#Bean
public JdbcMessageStore jdbcMessageStore() {
JdbcMessageStore jdbcMessageStore = new JdbcMessageStore(dataSource);
return jdbcMessageStore;
}
#Bean
public JobParametersToApiParametersTransformer jobParametersToApiParametersTransformer() {
return new JobParametersToApiParametersTransformer() {
#Override
protected ApiParameters transform(JobParameters jobParameters) {
ApiParameters.ApiParametersBuilder builder = ApiParameters.builder();
MultiValueMap<String, String> multiValueMap = new LinkedMultiValueMap<>();
String level = jobParameters.getString("level");
if (!StringUtils.isEmpty(level)) {
multiValueMap.set("level", level);
}
String fields = jobParameters.getString("fields");
if (!StringUtils.isEmpty(fields)) {
multiValueMap.set("fields", fields);
}
String filter = jobParameters.getString("filter");
if (filter != null) {
try {
JsonNode jsonNode = objectMapper.readTree(filter);
if (jsonNode != null && jsonNode.isArray()) {
List<ApiFilteringParameters> filteringParametersList = new ArrayList<>();
List<ApiSingleValueFilteringParameters> singleValueFilteringParameters = new ArrayList<>();
ArrayNode arrayNode = (ArrayNode) jsonNode;
arrayNode.forEach(node -> {
String field = node.get("field").asText();
String operator = node.get("operator").asText();
if (!StringUtils.isEmpty(field) && !StringUtils.isEmpty(operator)) {
String values = node.get("values").asText();
String[] valuesArray = !StringUtils.isEmpty(values) ? values.split(",") : null;
if (valuesArray != null) {
if (valuesArray.length > 1) {
filteringParametersList.add(ApiFilteringParameters
.of(field, Operator.valueOf(operator), valuesArray));
} else {
singleValueFilteringParameters.add(ApiSingleValueFilteringParameters
.of(field, Operator.valueOf(operator), valuesArray[0]));
}
}
}
});
if (!filteringParametersList.isEmpty()) {
builder.filterings(filteringParametersList);
}
if (!singleValueFilteringParameters.isEmpty()) {
builder.filterings(singleValueFilteringParameters);
}
}
} catch (IOException e) {
throw new UncheckedIOException(e);
}
}
String start = jobParameters.getString("time_ranges.start");
String end = jobParameters.getString("time_ranges.end");
String since = jobParameters.getString("time_range.since");
String until = jobParameters.getString("time_range.until");
if (!StringUtils.isEmpty(start) && !StringUtils.isEmpty(end)) {
builder.timeRanges(ApiParameters.timeRanges(start, end));
} else if (!StringUtils.isEmpty(since) && !StringUtils.isEmpty(until)) {
builder.timeRange(new TimeRange(since, until));
}
String actionBreakdowns = jobParameters.getString("action_breakdowns");
if (!StringUtils.isEmpty(actionBreakdowns)) {
multiValueMap.set("action_breakdowns", actionBreakdowns);
}
String attributionWindows = jobParameters.getString("action_attribution_windows");
if (attributionWindows != null) {
try {
multiValueMap
.set("action_attribution_windows",
objectMapper.writeValueAsString(attributionWindows.split(",")));
} catch (JsonProcessingException e) {
e.printStackTrace();
}
}
builder.multiValueMap(multiValueMap);
String pageSize = jobParameters.getString("pageSize");
if (!StringUtils.isEmpty(pageSize)) {
builder.limit(pageSize);
}
return builder.build();
}
};
}
}
Here is how message flows:
1. channel[async_ads_insights] ->IntegrationFlow[amqpOutboundAsyncAdsInsights]->[AMQP]->IntegrationFlow[amqpAdsInsightsAsyncJobRequestFlow]->channel[ad_report_run_polling_channel]->IntegrationFlow[adReportRunPollingLoopFlow]-IF END LOOP->channel[batch_launch_channel] ELSE -> channel[ad_report_run_polling_channel]
2. channel[batch_launch_channel] -> IntegrationFlow[amqpOutbound]-> IntegrationFlow[amqpLaunchSpringBatchJobFlow]
3. Spring Batch Job is launched.
The exception isn't thrown immediately after both instances are started, but after a while. Launching Spring Batch Jobs do succeeds but then start to fail with "A job instance already exists and is complete for..."
The job is for retrieving facebook ads results.
I would appreciate your insights into what is causing the error above.
I also have this configuration which does not use AMQP and works without any problem, but it is only for one instance.
#Configuration
#Conditional(SimpleBatchLaunchCondition.class)
#Slf4j
public class SimpleBatchLaunchIntegrationFlows {
#Autowired
SpringBatchLauncher batchLauncher;
#Autowired
DataSource dataSource;
#Bean(name = "batch_launch_channel")
public MessageChannel batchLaunchChannel() {
return MessageChannels.queue(jdbcChannelMessageStore(), "batch_launch_channel").get();
}
#Bean
public ChannelMessageStoreQueryProvider channelMessageStoreQueryProvider() {
return new MySqlChannelMessageStoreQueryProvider();
}
#Bean
public JdbcChannelMessageStore jdbcChannelMessageStore() {
JdbcChannelMessageStore channelMessageStore = new JdbcChannelMessageStore(dataSource);
channelMessageStore.setChannelMessageStoreQueryProvider(channelMessageStoreQueryProvider());
channelMessageStore.setUsingIdCache(true);
channelMessageStore.setPriorityEnabled(true);
return channelMessageStore;
}
#Bean
public IntegrationFlow launchSpringBatchJobFlow(#Qualifier("batch_launch_channel")
MessageChannel batchLaunchChannel) {
return IntegrationFlows.from(batchLaunchChannel)
.handle(message -> {
String jobName = (String) message.getHeaders().get("job_name");
JobParameters jobParameters = (JobParameters) message.getPayload();
batchLauncher.launchJob(jobName, jobParameters);
}, e->e.poller(Pollers.fixedRate(500).receiveTimeout(500))).get();
}
}

Answer 1

Refer to the Spring Batch documentation. When launching a new instance of a job, the job parameters must be unique.
A common solution is to add a dummy parameter with a UUID or similar but batch provides a strategy, e.g to increment a numeric parameter each time.
EDIT
There is a certain class of exceptions where the members of which are considered irrecoverable (fatal) and it makes no sense to attempt redelivery.
Examples include MessageConversionException - if we can't convert it the first time, we probably can't convert on a redelivery. The ConditionalRejectingErrorHandler is the mechanism by which we detect such exceptions, and cause them to be permanently rejected (and not redelivered).
Other exceptions cause the message to be redelivered by default - there is another property defaultRequeuRejected which can be set to false to permanently reject all failures (not recommended).
You can customize the error handler by subclassing its DefaultExceptionStrategy - override isUserCauseFatal(Throwable cause) to scan the cause tree to look for a JobInstanceAlreadyCompleteException and return true (cause.getCause().getCause() instanceof ...)
I think it was triggered by the error thrown by the "SpringBatch job running already" exception.
That still indicates you have somehow received a second message with the same parameters; it's a different error because the original job is still running; that message is rejected (and requeued) but on subsequent deliveries you get the already completed exception.
So, I still say the root cause of your problem is duplicate requests, but you can avoid the behavior with a customized error handler in the channel adapter's listener container.
I suggest you log the duplicate message so you can figure out why you are getting them.