+1, But I think we can apply the way of centralized processing for the issues of this type to improve the efficiency.
[ Full content available at: https://github.com/apache/flink/pull/6591 ] This message was relayed via gitbox.apache.org for [email protected]
