[
https://issues.apache.org/jira/browse/LUCENE-4583?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13655525#comment-13655525
]
Jack Krupansky commented on LUCENE-4583:
----------------------------------------
bq. abusing docvalues as stored fields
Great point. I have to admit that I still don't have a 100% handle on the use
case(s) for docvalues vs. stored fields, even though I've asked on the list. I
mean, sometimes the chatter seems to suggest that dv is the successor to stored
values. Hmmm... in that case, I should be able to store the full text of a 24
MB PDF file in a dv. Now, I know that isn't true.
Maybe we just need to start with some common use cases, based on size: tiny (16
bytes or less), small (256 or 1024 bytes or less), medium (up to 32K), and
large (upwards of 1MB, and larger.) It sounds like large implies stored field.
A related "concern" is dv or stored fields that need a bias towards being in
memory and in the heap, vs. a bias towards being "off heap". Maybe the size
category is the hint: tiny and small bias towards on-heap, medium and certainly
large bias towards off-heap. If people are only going towards DV because they
think they get off-heap, then maybe we need to reconsider the model of what DV
vs. stored is really all about. But then that leads back to DV somehow morphing
out of column-stride fields.
> StraightBytesDocValuesField fails if bytes > 32k
> ------------------------------------------------
>
> Key: LUCENE-4583
> URL: https://issues.apache.org/jira/browse/LUCENE-4583
> Project: Lucene - Core
> Issue Type: Bug
> Components: core/index
> Affects Versions: 4.0, 4.1, 5.0
> Reporter: David Smiley
> Priority: Critical
> Fix For: 4.4
>
> Attachments: LUCENE-4583.patch, LUCENE-4583.patch, LUCENE-4583.patch
>
>
> I didn't observe any limitations on the size of a bytes based DocValues field
> value in the docs. It appears that the limit is 32k, although I didn't get
> any friendly error telling me that was the limit. 32k is kind of small IMO;
> I suspect this limit is unintended and as such is a bug. The following
> test fails:
> {code:java}
> public void testBigDocValue() throws IOException {
> Directory dir = newDirectory();
> IndexWriter writer = new IndexWriter(dir, writerConfig(false));
> Document doc = new Document();
> BytesRef bytes = new BytesRef((4+4)*4097);//4096 works
> bytes.length = bytes.bytes.length;//byte data doesn't matter
> doc.add(new StraightBytesDocValuesField("dvField", bytes));
> writer.addDocument(doc);
> writer.commit();
> writer.close();
> DirectoryReader reader = DirectoryReader.open(dir);
> DocValues docValues = MultiDocValues.getDocValues(reader, "dvField");
> //FAILS IF BYTES IS BIG!
> docValues.getSource().getBytes(0, bytes);
> reader.close();
> dir.close();
> }
> {code}
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
