Spring/spring-integration
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

GH-9801: Kafka: generate ID & TIMESTAMP headers by default

Browse Source 
Fixes: https://github.com/spring-projects/spring-integration/issues/9801

Spring Integration Apache Kafka inbound channel adapters have always produced
messages without `ID` & `TIMESTAMP` headers.
See `MessagingMessageConverter` used over there by default.
However, for consistency with the rest of Spring Integration channel adapters,
it would be better to have Kafka-specific behave same way.

* Configure default `MessagingMessageConverter` in the `KafkaMessageSource` & `KafkaMessageDrivenChannelAdapter`
for generating `ID` & `TIMESTAMP` headers.
This commit is contained in:
Artem Bilan
2025-01-30 16:53:09 -05:00
parent c44363a69d
commit 054ff57728
5 changed files with 37 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
src/reference/antora/modules/ROOT/pages/whats-new.adoc
View File

@@ -27,27 +27,34 @@ The `AbstractCorrelatingMessageHandler` does not throw an `IllegalArgumentExcept
Instead, such a collection is wrapped into a single reply message.
See xref:aggregator.adoc[Aggregator] for more information.
[[x6.4-correlation-changes]]
[[x6.5-correlation-changes]]
== The `discardIndividuallyOnExpiry` Option For Correlation Handlers
The aggregator and resequencer can now discard the whole expired group as a single message via setting `discardIndividuallyOnExpiry` to `false`.
See xref:aggregator.adoc#releasestrategy[ReleaseStrategy] for more information.
[[x6.4-message-store-with-locks]]
[[x6.5-message-store-with-locks]]
== The `LockRegistry` in the `MessageStore`
The `AbstractMessageGroupStore` now can be configured with a `LockRegistry` to perform series of persistent operation atomically.
See xref:message-store.adoc#use-lock-registry[Use LockRegistry] for more information.
[[x6.4-observation-changes]]
[[x6.5-observation-changes]]
== Micrometer Observation Changes
The `SourcePollingChannelAdapter` endpoint now starts a `CONSUMER` kind observation for the received message.
The `MessageReceiverContext` now distinguishes between `handler`, `message-source` and `message-producer` values for the `spring.integration.type` low cardinality tag.
See xref:metrics.adoc#micrometer-observation[Micrometer Observation] for more information.
[[x6.4-mqtt-changes]]
[[x6.5-mqtt-changes]]
== Optional Paho MQTT Dependencies
The `org.eclipse.paho:org.eclipse.paho.client.mqttv3` dependency for `spring-integration-mqtt` is now also optional as `org.eclipse.paho:org.eclipse.paho.mqttv5.client` always was.
See xref:mqtt.adoc[MQTT Support] for more information.
See xref:mqtt.adoc[MQTT Support] for more information.
[[x6.5-kafka-changes]]
== Apache Kafka support Changes
The `KafkaMessageSource` and `KafkaMessageDrivenChannelAdapter` now generate `MessageHeaders.ID` and `MessageHeaders.TIMESTAMP` headers by default as the rest of Spring Integration channel adapters.
The behavior can be restored to the previous with injection of the `MessagingMessageConverter` with default settings.
See xref:kafka.adoc[Apache Kafka Support] for more information.