Yufei Zhang created FLINK-26159: ----------------------------------- Summary: Pulsar Connector: should add description MAX_FETCH_RECORD in doc to explain slow consumption Key: FLINK-26159 URL: https://issues.apache.org/jira/browse/FLINK-26159 Project: Flink Issue Type: Bug Components: Connectors / Pulsar Reporter: Yufei Zhang
Pulsar source connector can consume slow when the data volume is low (like 1 record/s), this is because the MAX_FETCH_RECORD is set to 100 by default, meaning that the pulsar source emits records either after a timeout (1s) or buffered messages has reached 100. Users will observe each message has a 1s delay. This can be add to the documentation and inform users how to change MAX_FETCH_RECORD -- This message was sent by Atlassian Jira (v8.20.1#820001)