Gabriel39 opened a new pull request, #66602: URL: https://github.com/apache/doris/pull/66602
## Proposed changes - keep whole Parquet files as FE scan tasks for File Scanner V2, including Iceberg and native Paimon paths - expand each file parent into row-group child tasks in BE after parsing the footer once - retain immutable Parquet metadata and table-format descriptors across concurrent child readers - share bounded reader-local cache state by physical file and correct cache/request accounting - harden dynamic split lifecycle, staged range ownership, specialized RPC accounting, and failure cleanup found while reviewing #66548 ## Validation - FE Checkstyle and 79 Iceberg/Paimon unit tests - BE ASAN unit-test binary build - 43 focused BE ASAN tests in the latest review pass, covering scanner scheduling, Parquet metadata and staged-file sharing, staged predicate/lazy-output activation, Paimon forwarding, file cache accounting, and buffered readers - clang-format 16 verification for every affected C/C++ source and header This PR is based on and intended to follow #66548. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
