lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ryan McKinley <ryan...@gmail.com>
Subject Re: Why isn't this working?
Date Fri, 28 Aug 2009 03:36:09 GMT

On Aug 27, 2009, at 10:35 PM, Paul Tomblin wrote:

> Yesterday or the day before, I asked specifically if I would need to
> restart the Solr server if somebody else loaded data into the Solr
> index using the EmbeddedServer, and I was told confidently that no,
> the Solr server would see the new data as soon as it was committed.
> So today I fired up the Solr server (and after making
> apache-tomcat-6.0.20/solr/data a symlink to where the Solr data really
> lives and restarting the web server), and did some queries.  Then I
> ran a program that loaded a bunch of data and committed it.  Then I
> did the queries again.  And the new data is NOT showing.  Using Luke,
> I can see 10022 documents in the index, but the Solr statistics page
> (http://localhost:8080/solrChunk/admin/stats.jsp) is still showing
> 8677, which is how many there were before I reloaded the data.
>
> So am I doing something wrong, or was the assurance I got yesterday
> that this is possible wrong?
>

did not follow the advice from yesterday... but...

the "commit" word can be a but misleading, it could also be called  
"reload"

Say you have an embedded solr server and an http solr server pointed  
to the same location.
1.  make sure only is read only!  otherwise you can make a mess.
2. calling commit on the embedded solr instance, will not have any  
effect on the http instance UNTIL you call commit (reload) on the http  
instance.

ryan

Mime
View raw message