[
https://issues.apache.org/jira/browse/LUCENE-7710?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mike Drob updated LUCENE-7710:
------------------------------
Attachment: LUCENE-7710.patch
Attaching simple patch to include more detail in the exception thrown. All
tests under lucene/core passed locally for me.
I think we should apply this to both trunk and branch_6x.
[~mikemccand] - Want to take a look? I feel like I'm playing whack-a-mole with
improving these exceptions individually as I find them in my logs.
> BlockPackedReader to throw better exception
> -------------------------------------------
>
> Key: LUCENE-7710
> URL: https://issues.apache.org/jira/browse/LUCENE-7710
> Project: Lucene - Core
> Issue Type: Improvement
> Affects Versions: 4.10.3
> Reporter: Mike Drob
> Attachments: LUCENE-7710.patch
>
>
> BlockPackedReader doesn't tell us which file we failed reading. Here's a
> stack trace from a 4.10.3 install, but it applies to trunk as well.
> {noformat}
> org.apache.solr.common.SolrException; null:java.io.IOException: Corrupted
> at
> org.apache.lucene.util.packed.BlockPackedReader.<init>(BlockPackedReader.java:56)
> at
> org.apache.lucene.codecs.lucene42.Lucene42DocValuesProducer.loadNumeric(Lucene42DocValuesProducer.java:204)
> at
> org.apache.lucene.codecs.lucene42.Lucene42DocValuesProducer.getNumeric(Lucene42DocValuesProducer.java:174)
> at
> org.apache.lucene.codecs.perfield.PerFieldDocValuesFormat$FieldsReader.getNumeric(PerFieldDocValuesFormat.java:248)
> at
> org.apache.lucene.index.SegmentCoreReaders.getNumericDocValues(SegmentCoreReaders.java:194)
> at
> org.apache.lucene.index.SegmentReader.getNumericDocValues(SegmentReader.java:229)
> at org.apache.lucene.search.FieldCacheImpl.getLongs(FieldCacheImpl.java:883)
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
