[
https://issues.apache.org/jira/browse/ORC-697?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17361578#comment-17361578
]
swolf commented on ORC-697:
---------------------------
Hi, [~sfiend] :
No, the problem remains unresolved. From the stripe's min/max index (min: 0,
max: 9268558), the data is normal, having no outliers. The file is compressed
by ZSTD, and the data buffer could be decompressed normally, so I think the
file is not corrupted. Having analyzed the RLEv2 encoding code and related
issues, I can't find a solution to the problem. It's a tricky problem,
[~omalley] could you help us? Thx.
> Improve Scan tool to report where files are corrupted.
> ------------------------------------------------------
>
> Key: ORC-697
> URL: https://issues.apache.org/jira/browse/ORC-697
> Project: ORC
> Issue Type: Improvement
> Components: tools
> Affects Versions: 1.7.0
> Reporter: Owen O'Malley
> Assignee: Owen O'Malley
> Priority: Major
> Fix For: 1.7.0
>
>
> We recently had a case where a bad machine was causing corruption in ORC
> files. In the process of debugging that, I extended the scan tool to report
> where the corruption was.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
