reactor-kafka无traceId

概览

springboot 2.6.7
reactor-kakfa 1.3.11
spring-cloud-starter-sleuth 3.1.2

解决

不止为何日志没有traceId

按照sleuth的文档，说这么写就行，实际没生效，不知道是否版本问题

@Bean("kafkaReceiver")
    public KafkaReceiver<String, String> kafkaReceiver(TracingKafkaConsumerFactory tracingKafkaConsumerFactory) {
        return KafkaReceiver.create(tracingKafkaConsumerFactory, options);
    }

最后在处理消息处，单独添加以下代码

import org.springframework.cloud.sleuth.Span;
import org.springframework.cloud.sleuth.SpanAndScope;
import org.springframework.cloud.sleuth.Tracer;
import org.springframework.cloud.sleuth.propagation.Propagator;

@PostConstruct
    public void handleMessage() {
        kafkaReceiver.receive()
                .flatMap(receiverRecord -> {
                    Span.Builder builder = propagator.extract(receiverRecord, extractor);
                    Span childSpan = builder.name("on-message").start();
                    SpanAndScope spanAndScope = new SpanAndScope(childSpan, tracer.withSpan(childSpan));

                    
                    return dohanle(receiverRecord)
                            .flatMap(valueResult -> {
                                receiverRecord.receiverOffset().acknowledge();
                                return Mono.just(receiverRecord);
                            })
                            .doFinally(signalType -> {
                                spanAndScope.close();
                            });
                })
                .subscribe();
    }

结果

traceId有了，kafka共用一个
spanId为每条消息单独，并且会向下传播