Re: [VuFind-Tech] Facets -- where are the bottlenecks?

[Alan R. <ala...@mn...>]

Hello,

I had this problem too, especially with topic facets as there are so
many. I added an option to the facet query in SOLR.php(now solr.php)
late last week and it seems to have helped. Have not had time to report
on it.

The original section in Solr.php starting with line 448 is:

// Build Facet Options
if ($facet) {
  $options['facet'] = 'true';
  $options['facet.mincount'] = 1;
  $options['facet.limit'] = (isset($facet['limit'])) ? $facet['limit'] :
null;
  $options['facet.field'] = (isset($facet['field'])) ? $facet['field'] :
null;
  $options['facet.prefix'] = (isset($facet['prefix'])) ?
$facet['prefix'] : null;
  $options['facet.sort'] = (isset($facet['sort'])) ? $facet['sort'] :
null;
}

I added the additional facet.enum.cache.minDf option after doing some
research:

// Build Facet Options
if ($facet) {
  $options['facet'] = 'true';
  $options['facet.mincount'] = 1;
  $options['facet.enum.cache.minDf'] = 1000;
  $options['facet.limit'] = (isset($facet['limit'])) ? $facet['limit'] :
null;
  $options['facet.field'] = (isset($facet['field'])) ? $facet['field'] :
null;
  $options['facet.prefix'] = (isset($facet['prefix'])) ?
$facet['prefix'] : null;
  $options['facet.sort'] = (isset($facet['sort'])) ? $facet['sort'] :
null;
}

give it a test -- al

On Mon, 2008-08-04 at 16:21 -0500, Andrew Nagy wrote:
> Solr processes all the work with the facets - all that VuFind does is take the data from solr and display it.  All the performance should happen with Jetty and Solr - mainly with Solr.  The best place to look is with the cache settings.  I spoke to some of the folks at CNET and said that the cache settings can be seen as almost black magic.  It's very hard to determine the best settings - but minor tweaks can have major differences.
> 
> Plus also your hardware infrastructure like - that could also make a major impact.  The more memory the faster the facets.
> 
> Hope this helps.
> 
> Andrew
> 
> > -----Original Message-----
> > From: vuf...@li... [mailto:vufind-tech-
> > bo...@li...] On Behalf Of Barnett, Jeffrey
> > Sent: Monday, August 04, 2008 4:57 PM
> > To: vuf...@li...
> > Subject: [VuFind-Tech] Facets -- where are the bottlenecks?
> >
> > Ever since we got our full 7.8million records loaded we've experienced
> > significant delays "Loading Narrow Options".  Following discussion here
> > and in solrmarc, I think we have the solr tuning options set to
> > reasonable values ... queries use less than 5% cpu. The one thing that
> > made some improvement was when we decided to drop one of the facet
> > categories that was confusing test users.  That leads me to think the
> > problem might be in the javascript and/or AJAX that is building the
> > facet panel.  Are there places that that might be tweaked or monitored?
> > Could we omit counting facets after the top 5?
> >
> > -----------------------------------------------------------------------
> > --
> > This SF.Net email is sponsored by the Moblin Your Move Developer's
> > challenge
> > Build the coolest Linux based applications with Moblin SDK & win great
> > prizes
> > Grand prize is a trip for two to an Open Source event anywhere in the
> > world
> > http://moblin-contest.org/redirect.php?banner_id=100&url=/
> > _______________________________________________
> > Vufind-tech mailing list
> > Vuf...@li...
> > https://lists.sourceforge.net/lists/listinfo/vufind-tech
> 
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Vufind-tech mailing list
> Vuf...@li...
> https://lists.sourceforge.net/lists/listinfo/vufind-tech
-- 
Alan Rykhus
PALS, A Program of the Minnesota State Colleges and Universities 
(507)389-1975
ala...@mn...