How to configure Jaeger Collector/Ingester to work with a cloud Kafka cluster.

[georgespingos]

Hi all,

the title says it all pretty much. I want to configure Jaeger Collector and Ingester to work with a could instance I have provisioned. If it helps at all, I am running this within a docker-compose stack with each individual Jaeger component configured separately as a service (not an all-in-one instance) and my cloud provider is Confluent could.

collector configuration (key/secret are obfuscated):

    image: jaegertracing/jaeger-collector
    container_name: jaeger-collector
    networks:
      - kafka-net
    ports:
      - "14250:14250"
      - "14267:14267"
      - "14268:14268" # HTTP collector port to receive spans
      - "14269:14269" # HTTP health check port
    restart: on-failure
    environment:
      LOG_LEVEL: "debug"
      SPAN_STORAGE_TYPE: "kafka"
      KAFKA_BROKERS: "pkc-lzoyy.europe-west6.gcp.confluent.cloud:9092"
      KAFKA_PRODUCER_BROKERS: "pkc-lzoyy.europe-west6.gcp.confluent.cloud:9092"
      KAFKA_PRODUCER_ENCODING: 'protobuf'
      KAFKA_PRODUCER_PLAINTEXT_PASSWORD: "JKSDHFJDHSPKSCSMLD+ZAM9UlXJ/Q5/tYl2NP6gu"
      KAFKA_PRODUCER_PLAINTEXT_USERNAME: "24YIUIUNKUKK5CK"
      KAFKA_PRODUCER_TOPIC: spans

I keep getting "error":"kafka: client has run out of available brokers to talk to (Is your cluster reachable?)" but I have confirmed that the Kafka cluster is reachable from inside the container.

Anyone ever tried anything similar or has any input it would be greatly appreciated.
~GS