Filr should migrate to the Solr indexing engine, it is massively faster, more scalable, and can index more content types.
by: Mark C. | over a year ago | Other
by: Mark C. | over a year ago | Other
Filr should migrate to the Solr indexing engine, it is massively faster, more scalable, and can index more content types.
by: Mark C. | over a year ago | Other
Comments
From what I read Solr is just a service sitting on top of the Lucene indexer; Solr is not the indexer itself. Looks like it has some nice benefits, not sure if that would benefit Filr directly too. Maybe if the search servers do the replication themselves vs the Filr Application servers doing that, this would offload the Filr Application servers a bit I would think. A lot Solr provides is probably done today in the Filr application server. But if it makes Filr and Vibe better you have my vote.
http://www.lucenetutorial.com/lucene-vs-solr.html
http://lucene.apache.org/solr/features.html
I support since indexing is a heavy load on the servers especially when we enable file content.
Or even Elasticsearch which might add benefit from its ability to scale and replicate. It seems impressively fast with a lot of the data we feed it from logs for analytics.
Hi All,
We have started working to first move to the latest version of Lucene. We expect this to boost up Filr's performance.
Note that Solr uses Lucene under the hood, Solr indexes and Lucene indexes are one and the same thing.
To summarize: we don't have immediate plans to move to Solr or Elastic search for Filr. Instead - we will first update and refine the integration we have with Lucene in Filr 4.0