vinothchandar commented on issue #18135: URL: https://github.com/apache/hudi/issues/18135#issuecomment-4071233690
@vamshipasunuru1 @shangxinli @nbalajee Got around to thinking about this, now that we are further along. I agree that the core change needed is > Return null for Hudi meta columns (_hoodie_record_key, _hoodie_commit_time, _hoodie_partition_path, _hoodie_file_name, _hoodie_commit_seqno) if the parquet file does not have columns. On the writer side, I wonder if we can simply get away by writing the `FILES` partition metadata (bootstrap etc was designed pre-metadata table), so all queries start seeing old partitions and their files? If a file group id is required, we can extend the metadata to store a file group id that we can generate one time during population of FILES metadata. cc @nsivabalan @bhasudha here again.. to help take this forward. @prashantwason what do you think? -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
