On Tue, Jul 30, 2013 at 11:48 PM, Robert Stewart <robert_stew...@epam.com>wrote:
> Also we need to issue frequent commits since we are constantly streaming > new content into the system. I'd like to say show me profiler snapshot, but after that note. Solr's filter/field caches are top level datastructures, hence they are blown off on every commit. Therefore even if you solve all other challenges, top-level reader is a bummer. >From what I've heard Lucene Facets and ElasticSearch should be more suitable for frequent commits. Technically what are you saying is somehow achived by http://wiki.apache.org/solr/SimpleFacetParameters#facet.enum.cache.minDf,but after commit cache says, bye-bye. -- Sincerely yours Mikhail Khludnev Principal Engineer, Grid Dynamics <http://www.griddynamics.com> <mkhlud...@griddynamics.com>
