yihua opened a new issue, #20092: URL: https://github.com/apache/hudi/issues/20092
Metadata table lookups that run in Spark tasks ship far more than they read. The key prefix and sharded lookups in `HoodieBackedTableMetadata` (column stats, expression, record and secondary index) capture the whole table metadata, including both meta clients and the file system view, and a task can list the data table timeline and read rollback metadata to compute the valid instants. The record index and metadata bloom filter functions on the write path ship the `HoodieTable` and open a new metadata reader in every task. The column stats read ships every candidate file name in every task. Proposal: resolve a small serializable reader for one metadata table partition on the driver, ship or broadcast only that reader to the tasks, and broadcast the candidate file names. part of #20064 -- 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]
