You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "grzegorz8 (via GitHub)" <gi...@apache.org> on 2026/03/10 09:30:07 UTC

Re: [PR] [FLINK-32380] Support Java Records with PojoTypeInfo/Serializer [flink]

grzegorz8 commented on PR #23490:
URL: https://github.com/apache/flink/pull/23490#issuecomment-4029907165

   @gyfora Hey! Are these benchmarks available somewhere? I was not able to find them on my own.
   
   ```
   RecordSerializationBenchmark.copyPojo         thrpt   30  6825.981 ± 32.221  ops/ms
   RecordSerializationBenchmark.copyRecord       thrpt   30  5949.899 ± 94.196  ops/ms
   RecordSerializationBenchmark.copyPojoReuse    thrpt   30  6757.649 ± 91.903  ops/ms
   RecordSerializationBenchmark.copyRecordReuse  thrpt   30  5974.339 ± 51.605  ops/ms
   RecordSerializationBenchmark.copyPojoView     thrpt   30   929.945 ±  3.241  ops/ms
   RecordSerializationBenchmark.copyRecordView   thrpt   30   916.406 ± 53.823  ops/ms
   RecordSerializationBenchmark.readPojo         thrpt   30  1070.375 ±  5.260  ops/ms
   RecordSerializationBenchmark.readRecord       thrpt   30  1041.910 ± 11.632  ops/ms
   RecordSerializationBenchmark.writePojo        thrpt   30  1225.184 ±  3.824  ops/ms
   RecordSerializationBenchmark.writeRecord      thrpt   30  1234.543 ±  4.738  ops/ms
   ```
   
   I did my own simple benchmarks and it seems that for objects with higher number of fields, deserialization of Record is noticeably faster than deserialization of Pojo. Pojo deserialization, in turn is faster or equally fast only for a small objects.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org