[
https://issues.apache.org/jira/browse/SOLR-17449?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Seunghan Jung updated SOLR-17449:
---------------------------------
Description:
For fields of type `bboxField`, derived fields such as `minX`, `maxX`, `minY`,
`maxY`, etc., are added to the schema and indexed.
However, this causes issues during atomic updates. During an atomic update,
when a new document is moved and indexed, the fields of type `bboxField` are
re-indexed with `minX`, `maxX`, `minY`, `maxY` just as they were initially.
Since the original document already contains these fields, they are indexed
again in the new document. However, because they are already indexed by the
`bbox` field, this results in duplicate indexing. If the `numberType` attribute
of the bbox field type has `docValues=true`, an error occurs due to the
docValues being written twice.
Here is the error message for this case:
{code:java}
Caused by: java.lang.IllegalArgumentException: DocValuesField "bbox__maxX"
appears more than once in this document (only one value is allowed per field)
at
org.apache.lucene.index.NumericDocValuesWriter.addValue(NumericDocValuesWriter.java:53)
~[?:?]
at
org.apache.lucene.index.IndexingChain.indexDocValue(IndexingChain.java:937)
~[?:?]
at
org.apache.lucene.index.IndexingChain.processField(IndexingChain.java:723)
~[?:?]
at
org.apache.lucene.index.IndexingChain.processDocument(IndexingChain.java:576)
~[?:?]
at
org.apache.lucene.index.DocumentsWriterPerThread.updateDocuments(DocumentsWriterPerThread.java:242)
~[?:?]
at
org.apache.lucene.index.DocumentsWriter.updateDocuments(DocumentsWriter.java:432)
~[?:?]
at
org.apache.lucene.index.IndexWriter.updateDocuments(IndexWriter.java:1545)
~[?:?]
at
org.apache.lucene.index.IndexWriter.updateDocuments(IndexWriter.java:1521)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.updateDocOrDocValues(DirectUpdateHandler2.java:1062)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.doNormalUpdate(DirectUpdateHandler2.java:421)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.addDoc0(DirectUpdateHandler2.java:374)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.addDoc(DirectUpdateHandler2.java:311)
~[?:?]{code}
Of course, setting the docValues="false" for the field type used in numberType
resolves the issue.
However, this is not explained in the[ Solr Reference
Guide|https://solr.apache.org/guide/solr/latest/query-guide/spatial-search.html#bboxfield].
Instead, the example schema shows docValues="true", which makes it seem like
this is how it should be configured.
!image-2024-09-13-17-45-17-161.png|width=1037,height=233!
was:
For fields of type `bboxField`, derived fields such as `minX`, `maxX`, `minY`,
`maxY`, etc., are added to the schema and indexed.
However, this causes issues during atomic updates. During an atomic update,
when a new document is moved and indexed, the fields of type `bboxField` are
re-indexed with `minX`, `maxX`, `minY`, `maxY` just as they were initially.
Since the original document already contains these fields, they are indexed
again in the new document. However, because they are already indexed by the
`bbox` field, this results in duplicate indexing. If the `numberType` attribute
of the bbox field type has `docValues=true`, an error occurs due to the
docValues being written twice.
Here is the error message for this case:
{code:java}
Caused by: java.lang.IllegalArgumentException: DocValuesField "bbox__maxX"
appears more than once in this document (only one value is allowed per field)
at
org.apache.lucene.index.NumericDocValuesWriter.addValue(NumericDocValuesWriter.java:53)
~[?:?]
at
org.apache.lucene.index.IndexingChain.indexDocValue(IndexingChain.java:937)
~[?:?]
at
org.apache.lucene.index.IndexingChain.processField(IndexingChain.java:723)
~[?:?]
at
org.apache.lucene.index.IndexingChain.processDocument(IndexingChain.java:576)
~[?:?]
at
org.apache.lucene.index.DocumentsWriterPerThread.updateDocuments(DocumentsWriterPerThread.java:242)
~[?:?]
at
org.apache.lucene.index.DocumentsWriter.updateDocuments(DocumentsWriter.java:432)
~[?:?]
at
org.apache.lucene.index.IndexWriter.updateDocuments(IndexWriter.java:1545)
~[?:?]
at
org.apache.lucene.index.IndexWriter.updateDocuments(IndexWriter.java:1521)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.updateDocOrDocValues(DirectUpdateHandler2.java:1062)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.doNormalUpdate(DirectUpdateHandler2.java:421)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.addDoc0(DirectUpdateHandler2.java:374)
~[?:?]
at
org.apache.solr.update.DirectUpdateHandler2.addDoc(DirectUpdateHandler2.java:311)
~[?:?]{code}
> bboxField subfield Error in atomic updating
> -------------------------------------------
>
> Key: SOLR-17449
> URL: https://issues.apache.org/jira/browse/SOLR-17449
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Components: Schema and Analysis
> Affects Versions: main (10.0)
> Reporter: Seunghan Jung
> Priority: Minor
> Attachments: image-2024-09-13-17-45-17-161.png
>
>
> For fields of type `bboxField`, derived fields such as `minX`, `maxX`,
> `minY`, `maxY`, etc., are added to the schema and indexed.
> However, this causes issues during atomic updates. During an atomic update,
> when a new document is moved and indexed, the fields of type `bboxField` are
> re-indexed with `minX`, `maxX`, `minY`, `maxY` just as they were initially.
> Since the original document already contains these fields, they are indexed
> again in the new document. However, because they are already indexed by the
> `bbox` field, this results in duplicate indexing. If the `numberType`
> attribute of the bbox field type has `docValues=true`, an error occurs due to
> the docValues being written twice.
> Here is the error message for this case:
> {code:java}
> Caused by: java.lang.IllegalArgumentException: DocValuesField "bbox__maxX"
> appears more than once in this document (only one value is allowed per field)
> at
> org.apache.lucene.index.NumericDocValuesWriter.addValue(NumericDocValuesWriter.java:53)
> ~[?:?]
> at
> org.apache.lucene.index.IndexingChain.indexDocValue(IndexingChain.java:937)
> ~[?:?]
> at
> org.apache.lucene.index.IndexingChain.processField(IndexingChain.java:723)
> ~[?:?]
> at
> org.apache.lucene.index.IndexingChain.processDocument(IndexingChain.java:576)
> ~[?:?]
> at
> org.apache.lucene.index.DocumentsWriterPerThread.updateDocuments(DocumentsWriterPerThread.java:242)
> ~[?:?]
> at
> org.apache.lucene.index.DocumentsWriter.updateDocuments(DocumentsWriter.java:432)
> ~[?:?]
> at
> org.apache.lucene.index.IndexWriter.updateDocuments(IndexWriter.java:1545)
> ~[?:?]
> at
> org.apache.lucene.index.IndexWriter.updateDocuments(IndexWriter.java:1521)
> ~[?:?]
> at
> org.apache.solr.update.DirectUpdateHandler2.updateDocOrDocValues(DirectUpdateHandler2.java:1062)
> ~[?:?]
> at
> org.apache.solr.update.DirectUpdateHandler2.doNormalUpdate(DirectUpdateHandler2.java:421)
> ~[?:?]
> at
> org.apache.solr.update.DirectUpdateHandler2.addDoc0(DirectUpdateHandler2.java:374)
> ~[?:?]
> at
> org.apache.solr.update.DirectUpdateHandler2.addDoc(DirectUpdateHandler2.java:311)
> ~[?:?]{code}
>
> Of course, setting the docValues="false" for the field type used in
> numberType resolves the issue.
> However, this is not explained in the[ Solr Reference
> Guide|https://solr.apache.org/guide/solr/latest/query-guide/spatial-search.html#bboxfield].
> Instead, the example schema shows docValues="true", which makes it seem like
> this is how it should be configured.
> !image-2024-09-13-17-45-17-161.png|width=1037,height=233!
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
