jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Marcel Reutegger <marcel.reuteg...@gmx.net>
Subject Re: Writing indexes takes a long time
Date Mon, 30 Nov 2009 13:54:28 GMT
Hi,

On Tue, Nov 24, 2009 at 17:23, pkrishnaswami
<prabhakar.krishnaswami@ge.com> wrote:
>
> Here is the section in repository.xml:
>
>        <SearchIndex
>                class="org.apache.jackrabbit.core.query.lucene.SearchIndex">
>
>
>
>
>
>        </SearchIndex>

unfortunately that didn't make it through either...

> Does the above entry in the repository.xml automatically index the payload
> which is stored as an array of bytes? If we want to disbale indexing of
> payload, should we specifically set a rule in the IndexConfiguration file. I
> appreaciate the response.

yes you should. if none of the index rules applies, then all
properties are indexed. If you don't want that, then you need to
define a rule that does not contain the jcr:data property.

regards
 marcel

> The version of jackRabbit being used: 1.5.4 (jackrabbit-core)
>
>
>
>
> Marcel Reutegger wrote:
>>
>> Hi,
>>
>> On Sat, Nov 21, 2009 at 01:10, pkrishnaswami
>> <prabhakar.krishnaswami@ge.com> wrote:
>>>
>>> We have observed that writing out JackRabbit indexes takes an unduly long
>>> time and we suspect this could be because of payload indexing. I would
>>> like
>>> some clarity on this. Here is the scenario:
>>>
>>>
>>> We save custom Nodes in our repository. Our custom Node (call A) extends
>>> nt:file and has many properties. In addition,we also make this custom
>>> node a
>>> parent to another custom node (call B) that extends nt:resource. We have
>>> an
>>> indexConfiguration file that specifically defines which properties in
>>> node
>>> "A" will be indexed. Custom Node B only contains a binary representation
>>> of
>>> XML or plain text contents (let call it payload). We specifically omitted
>>> an
>>> index-rule for custom Node B since we did not want the payload to be
>>> indexed, Would this disable indexing of payload?
>>
>> no, it doesn't. all properties are indexed when there is no applicable
>> rule.
>>
>>> I was reading up some documentation  of Nabble forum and there was a
>>> comment
>>> in one of the posts that the default configuration disables full-text
>>> indexing. What does this mean? It goes on to say that if full-text
>>> indexing
>>> is required, then we have to configure a SearchIndex section in
>>> repository.xml along with textFilterClasses property pointing to
>>> different
>>> extractors. I looked up the repository.xml we were using and sure enough
>>> there was a section like this:
>>>
>>>        <SearchIndex
>>>
>>>  class="org.apache.jackrabbit.core.query.lucene.SearchIndex">
>>>
>>>
>>>
>>>
>>>
>>>        </SearchIndex>
>>>
>>> Does this activate payload indexing? If it does, does it override the
>>> indexConfiguration file? If we didn't want payload indexing, should I
>>> comment or remove the param related textFilterClasses?
>>
>> the configuration did not make it though to the mailing list. can you
>> please post it again?
>>
>> please also state which version of jackrabbit you are using.
>>
>> regards
>>  marcel
>>
>>> --
>>> View this message in context:
>>> http://n4.nabble.com/Writing-indexes-takes-a-long-time-tp633631p633631.html
>>> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
>>>
>>
>>
>
> --
> View this message in context: http://n4.nabble.com/Writing-indexes-takes-a-long-time-tp633631p786947.html
> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
>

Mime
View raw message