lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Erick Erickson <>
Subject Re: Drawbacks of using Docvalues Fields
Date Wed, 01 Jun 2016 16:17:11 GMT
You can tell very little about performance with 50 documents,
so I wouldn't trust these results at all. In particular I'm pretty
sure that your search speed will suffer _greatly_ as you get
more and more documents in your corpus if you use only
DocValues but don't have indexed="true" set.


On Tue, May 31, 2016 at 10:51 PM, Chitra R <> wrote:
> Hi,
>       I have one doubt. Actually I have indexed 50 documents for both
> SortedNumericDocValuesField and IntField separately . Each document
> consists of 49 fields. After that, i performed indexing & searching ( with
> & without sorting) .
>      It seems DocValuesField indexes and searches the documents faster than
> normal FieldType (eg: IntField) and also it retrieves the values more
> efficent than normal field.  Is their any drawbacks to
> use SortedNumericDocValuesField over IntField?
> Also It seems like we can use Docvalues Field(String or Numeric)
> efficiently for any functionality(Searching and sorting). So generally, do
> we have any drawbacks of using DocValues Field(String and Numeric)?
> Known Drawbacks(on StringFields):
> -->docvalues don't work on analyzed fields
> -->search_analyzer cannot be specified on non-analyzed fields.
> is there any specific drawback on numeric docvalue fields? is range query
> not possible?? or anyother?
> Thanks,
> Chitra

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message