Get hands-on experience with 20+ free Google Cloud products and $300 in free credit for new customers.

ApacheBeamRunJavaPipeline composer/airflow Operator running Kafka source connection from airflow.

I am trying to run a dataflow java job that runs perfectly fine on dataflow runner when submitted without composer. The same job when tried from composer using dataflowrunner, some how composer is executing Kafka connection in airflow worker host rather than running from dataflow worker host. we only have firewall open from dataflow worker host and currently getting error fetching topic meta data timeout exception in composer.

java.lang.RuntimeException: org.apache.kafka.common.errors.TimeoutException: Timeout expired while fetching topic metadata

Is there any way to bypass jar execution on the airflow worker and only run the job from dataflow workers?

0 0 321
0 REPLIES 0
Top Labels in this Space
Top Solution Authors