You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-dev@lucene.apache.org by anuvenk <an...@hotmail.com> on 2007/11/29 22:05:23 UTC
solr version upgrade
should i re-index all the data if i upgrade to the larest solr version?
--
View this message in context: http://www.nabble.com/solr-version-upgrade-tf4899944.html#a14034971
Sent from the Solr - Dev mailing list archive at Nabble.com.
Re: solr version upgrade
Posted by Yonik Seeley <yo...@apache.org>.
On Nov 30, 2007 3:10 PM, anuvenk <an...@hotmail.com> wrote:
> Ok. Here is the parsedquery_toString
> <str name="parsedquery_toString">
> +(((text:need^0.8 | name:need^2.0)~0.01 (text:roommat^0.8 |
> name:roommat^2.0)~0.01 (text:agreement^0.8 | name:agreement^2.0)~0.01)~3)
> (text:"need roommat agreement"~50^0.8 | name:"need roommat
> agreement"~50^2.0)~0.01
> </str>
> The index was built a couple of months back.
> The stopword 'need' was recently added, a couple of days back.
> But you can see that its not getting removed at all before finding matches.
Schema changes don't take effect until Solr is restarted... was it?
If so, make sure there were no exceptions on startup (perhaps the file
couldn't be found or read?), or perhaps Solr is finding the wrong
version of the file... look around on your system for other files with
the same name in the classpath.
-Yonik
Re: solr version upgrade
Posted by anuvenk <an...@hotmail.com>.
Ok. Here is the parsedquery_toString
<str name="parsedquery_toString">
+(((text:need^0.8 | name:need^2.0)~0.01 (text:roommat^0.8 |
name:roommat^2.0)~0.01 (text:agreement^0.8 | name:agreement^2.0)~0.01)~3)
(text:"need roommat agreement"~50^0.8 | name:"need roommat
agreement"~50^2.0)~0.01
</str>
The index was built a couple of months back.
The stopword 'need' was recently added, a couple of days back.
But you can see that its not getting removed at all before finding matches.
Whereas i have a local test version on my PC, the same version of solr as my
work's.
Same schema, config. Index was built at the same time. Its the same as my
work's.
I added the stopword 'need' at the same time (couple of days back). But the
parsedquery_toString is different on mine:
<str name="parsedquery_toString">
+(((text:roommat^0.8 | name:roommat^2.0)~0.01 (text:agreement^0.8 |
name:agreement^2.0)~0.01)~3) (text:"roommat agreement"~50^0.8 |
name:"roommat agreement"~50^2.0)~0.01
</str>
And i get some results on my version while i get zero results in my work
version.
In the 'text' field definition in the schema, both index and query gets
passed through the stopwordfilter.
Why this discrepancy? I'm re-indexing the data on my work version. But can't
re-index the data everytime i add a new stopword right.
Yonik Seeley wrote:
>
> On Nov 29, 2007 6:26 PM, anuvenk <an...@hotmail.com> wrote:
>> The schema hasn't been updated in a long time though. I have the same
>> version
>> of solr & the same index & configs on my local PC and that seems to
>> exlude
>> the stop words i added while searching. Its quite puzzling why the new
>> stopwords wouldn't be taken in to account.
>
> It depends what you mean by "exlude the stop words i added while
> searching".
> If you have added those stopwords, they should be removed at query
> time by any analyzer using the stop filter. Those words will remain
> in the index until you reindex of course.
>
> Perhaps you could show the output of debugQuery=on for a sample query
> and how that differs from what you expect?
>
> -Yonik
>
>
--
View this message in context: http://www.nabble.com/solr-version-upgrade-tf4899944.html#a14092131
Sent from the Solr - Dev mailing list archive at Nabble.com.
Re: solr version upgrade
Posted by anuvenk <an...@hotmail.com>.
Ok. Here is the parsedquery_toString
<str name="parsedquery_toString">
+(((text:need^0.8 | name:need^2.0)~0.01 (text:roommat^0.8 |
name:roommat^2.0)~0.01 (text:agreement^0.8 | name:agreement^2.0)~0.01)~3)
(text:"need roommat agreement"~50^0.8 | name:"need roommat
agreement"~50^2.0)~0.01
</str>
The index was built a couple of months back.
The stopword 'need' was recently added, a couple of days back.
But you can see that its not getting removed at all before finding matches.
Whereas i have a local test version on my PC, the same version of solr as my
work's.
Same schema, config. Index was built at the same time. Its the same as my
work's.
I added the stopword 'need' at the same time (couple of days back). But the
parsedquery_toString is different on mine:
<str name="parsedquery_toString">
+(((text:roommat^0.8 | name:roommat^2.0)~0.01 (text:agreement^0.8 |
name:agreement^2.0)~0.01)~3) (text:"roommat agreement"~50^0.8 |
name:"roommat agreement"~50^2.0)~0.01
</str>
And i get some results on my version while i get zero results in my work
version.
In the 'text' field definition in the schema, both index and query gets
passed through the stopwordfilter.
Why this discrepancy? I'm re-indexing the data on my work version. But can't
re-index the data everytime i add a new stopword right.
Yonik Seeley wrote:
>
> On Nov 29, 2007 6:26 PM, anuvenk <an...@hotmail.com> wrote:
>> The schema hasn't been updated in a long time though. I have the same
>> version
>> of solr & the same index & configs on my local PC and that seems to
>> exlude
>> the stop words i added while searching. Its quite puzzling why the new
>> stopwords wouldn't be taken in to account.
>
> It depends what you mean by "exlude the stop words i added while
> searching".
> If you have added those stopwords, they should be removed at query
> time by any analyzer using the stop filter. Those words will remain
> in the index until you reindex of course.
>
> Perhaps you could show the output of debugQuery=on for a sample query
> and how that differs from what you expect?
>
> -Yonik
>
>
--
View this message in context: http://www.nabble.com/solr-version-upgrade-tf4899944.html#a14092131
Sent from the Solr - Dev mailing list archive at Nabble.com.
Re: solr version upgrade
Posted by Yonik Seeley <yo...@apache.org>.
On Nov 29, 2007 6:26 PM, anuvenk <an...@hotmail.com> wrote:
> The schema hasn't been updated in a long time though. I have the same version
> of solr & the same index & configs on my local PC and that seems to exlude
> the stop words i added while searching. Its quite puzzling why the new
> stopwords wouldn't be taken in to account.
It depends what you mean by "exlude the stop words i added while searching".
If you have added those stopwords, they should be removed at query
time by any analyzer using the stop filter. Those words will remain
in the index until you reindex of course.
Perhaps you could show the output of debugQuery=on for a sample query
and how that differs from what you expect?
-Yonik
Re: solr version upgrade
Posted by anuvenk <an...@hotmail.com>.
The schema hasn't been updated in a long time though. I have the same version
of solr & the same index & configs on my local PC and that seems to exlude
the stop words i added while searching. Its quite puzzling why the new
stopwords wouldn't be taken in to account.
Yonik Seeley wrote:
>
> On Nov 29, 2007 6:08 PM, anuvenk <an...@hotmail.com> wrote:
>> I got this doubt because i added a few stopwords and synonyms but even
>> after
>> restarting solr & tomacat, the stopwords and synonyms were still not
>> being
>> taken in to account by solr. Since we upgraded recently i wasn't sure if
>> that was the reason. Any idea why this could be happening?
>
> Changes to the schema will often require re-indexing.
>
> -Yonik
>
>
--
View this message in context: http://www.nabble.com/solr-version-upgrade-tf4899944.html#a14037532
Sent from the Solr - Dev mailing list archive at Nabble.com.
Re: solr version upgrade
Posted by Yonik Seeley <yo...@apache.org>.
On Nov 29, 2007 6:08 PM, anuvenk <an...@hotmail.com> wrote:
> I got this doubt because i added a few stopwords and synonyms but even after
> restarting solr & tomacat, the stopwords and synonyms were still not being
> taken in to account by solr. Since we upgraded recently i wasn't sure if
> that was the reason. Any idea why this could be happening?
Changes to the schema will often require re-indexing.
-Yonik
Re: solr version upgrade
Posted by anuvenk <an...@hotmail.com>.
I got this doubt because i added a few stopwords and synonyms but even after
restarting solr & tomacat, the stopwords and synonyms were still not being
taken in to account by solr. Since we upgraded recently i wasn't sure if
that was the reason. Any idea why this could be happening?
Yonik Seeley wrote:
>
> On Nov 29, 2007 4:05 PM, anuvenk <an...@hotmail.com> wrote:
>> should i re-index all the data if i upgrade to the larest solr version?
>
> No, it shouldn't be needed.
>
> -Yonik
>
>
--
View this message in context: http://www.nabble.com/solr-version-upgrade-tf4899944.html#a14037208
Sent from the Solr - Dev mailing list archive at Nabble.com.
Re: solr version upgrade
Posted by Yonik Seeley <yo...@apache.org>.
On Nov 29, 2007 4:05 PM, anuvenk <an...@hotmail.com> wrote:
> should i re-index all the data if i upgrade to the larest solr version?
No, it shouldn't be needed.
-Yonik