You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@flink.apache.org by "Yingjie Cao (Jira)" <ji...@apache.org> on 2020/09/21 09:53:00 UTC

[jira] [Created] (FLINK-19323) Small optimization of network layer record serialization

Yingjie Cao created FLINK-19323:
-----------------------------------

             Summary: Small optimization of network layer record serialization
                 Key: FLINK-19323
                 URL: https://issues.apache.org/jira/browse/FLINK-19323
             Project: Flink
          Issue Type: Sub-task
          Components: Runtime / Network
            Reporter: Yingjie Cao
             Fix For: 1.12.0


Currently, when serializing a record, the SpanningRecordSerializer will first skip 4 bytes for length and serialize the record. Then it gets the serialized record length and skip back to position 0 and write the length field. After that, it skip again to the tail of the serialized data. In fact, the last two skip can be avoid by writing length field to position 0 directly.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)