rangareddy commented on issue #16699:
URL: https://github.com/apache/hudi/issues/16699#issuecomment-5434698892
This issue was reviewed as part of the JIRA-migrated backlog triage
(HUDI-8605).
**Findings: the linked PR has not merged; keeping this open.**
PR #12393 (`[HUDI-8605] Fix Trino failure when reading corrupted block at
end of log file`) is **open and unmerged**.
Worth recording context that narrows the fix: generic corrupt-block handling
already exists in
`hudi-common/src/main/java/org/apache/hudi/common/table/log/HoodieLogFileReader.java`
- `:131-138` create a corrupt block by scanning for the next `MAGIC` marker or
EOF, `:243` logs `"Log {} has a corrupted block at {}"`, and `:246-249` rewind
and read the corrupted bytes up to `nextBlockOffset`.
So this is not missing corrupt-block handling; it is the specific
*end-of-file* case, where there is no following `MAGIC` marker to scan to,
surfacing through Trino. That is a narrower problem than the title suggests.
Also relevant now that the Trino connector is in-repo (`hudi-trino/`,
RFC-105 #18837): a regression test for this can live alongside the connector
rather than being verified out-of-tree.
--
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]
