[
https://issues.apache.org/jira/browse/HADOOP-20060?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18138455#comment-18138455
]
ASF GitHub Bot commented on HADOOP-20060:
-----------------------------------------
LuciferYang opened a new pull request, #8923:
URL: https://github.com/apache/hadoop/pull/8923
### Description of PR
JIRA: https://issues.apache.org/jira/browse/HADOOP-20060
When a `.crc` file exists but its header cannot be read (zero length,
truncated or bad magic), `ChecksumFSInputChecker` logs a warning and reads the
data without verification. `readVectored()` only hands the ranges to the data
stream when `sums == null`, which is not the case here, so it reads checksum
ranges from the broken file with `bytesPerSum` = 1 and fails with
`EOFException: Read past End of File`.
`hasCapability(VECTOREDIO_BUFFERS_SLICED)` also returns true. A zero-length
`.crc` is what a writer leaves when it dies before closing, and Parquet 1.16
uses vectored reads by default.
This patch makes `dataStreamToHandleVectorIO()` also delegate when checksums
are not being verified (`!needChecksum()`), as its javadoc already describes.
`ChecksumFileSystem#open` only builds this checker when checksum verification
is on, so in practice the new branch is taken only when the `.crc` header could
not be read.
### How was this patch tested?
Added
`TestLocalFSContractVectoredRead#testVectoredReadWithCorruptChecksumHeader` (4
bytes of garbage) and `#testVectoredReadWithEmptyChecksumFile` (zero-length
`.crc`). Both read the file normally, then check that the stream does not
report sliced vectored buffers and that `readVectored` returns the right data.
Without the fix both fail on the capability check, and the vectored read fails
with `EOFException`.
### For code changes:
- [x] Does the title of this PR start with the corresponding JIRA issue id
(e.g. 'HADOOP-17799. Your PR title ...')?
- [ ] Object storage: Have the integration tests been executed and the
endpoint declared according to the connector-specific documentation?
- [ ] If adding new dependencies to the code, are these dependencies
licensed in a way that is compatible for inclusion under [ASF
2.0](http://www.apache.org/legal/resolved.html#category-a)?
- [ ] If applicable, have you updated the `LICENSE`, `LICENSE-binary`,
`NOTICE-binary` files?
### AI Tooling
- [x] The PR includes the phrase "Contains content generated by Claude Opus".
- [x] My use of AI contributions follows the ASF legal policy
https://www.apache.org/legal/generative-tooling.html
Contains content generated by Claude Opus.
> ChecksumFileSystem: readVectored() fails when the .crc header is unreadable,
> while read() falls back to unverified reads
> ------------------------------------------------------------------------------------------------------------------------
>
> Key: HADOOP-20060
> URL: https://issues.apache.org/jira/browse/HADOOP-20060
> Project: Hadoop Common
> Issue Type: Bug
> Components: fs
> Reporter: Yang Jie
> Priority: Minor
>
> When a file's .crc exists but its header cannot be read (zero length,
> truncated, or bad magic), ChecksumFSInputChecker logs "Problem opening
> checksum file ... Ignoring exception" and reads the data without
> verification. readVectored() does not do the same. It only hands the ranges
> to the data stream when {{sums == null}}, and here {{sums}} is still open, so
> it computes checksum ranges with the default {{bytesPerSum}} of 1 and reads
> them from the broken .crc. That fails with "EOFException: Read past End of
> File", or with a ChecksumException at meaningless offsets if the .crc is
> large. hasCapability(VECTOREDIO_BUFFERS_SLICED) also returns true.
> A zero-length .crc is what a writer leaves behind when it dies before closing
> the file, because the header stays in a buffer until roughly 511 KiB of data
> has been written. Parquet 1.16 enables vectored reads by default, so Spark
> reading such a Parquet file from file:// reads the footer and then fails on
> the row groups. With Parquet 1.15 the same file can be read.
> The fix makes {{dataStreamToHandleVectorIO()}} also delegate when checksums
> are not being verified ({{!needChecksum()}}), as its javadoc already says it
> should.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
