You are viewing a plain text version of this content. The canonical link for it is here.

Posted to solr-user@lucene.apache.org by Jeetendra Gangele <je...@insideview.com> on 2015/04/28 17:48:53 UTC

Support of solr in Spark

Hi All

I have around 20 million company name and I want to index them.
Currently What I am doing I am tokenizing and for each token I am applying Metaphone 3 and then Stroring each token in Hbase.
When I get new query(company to match) I will again tokenize and apply metaphone3 as I did when I stored them in Hbase
Now for each token I will query Hbase and collate the result.

This seems inefficient and has some issue even after implementing the functionality of WordDelimiterFilterFactory<http://stackoverflow.com/questions/17707733/worddelimiterfilterfactory-not-including-all-permutations> and singleFilter factory.

I am thinking to index these companies name in solr since all the functionality already there?

Do we have support for spark?

Re: Support of solr in Spark

Posted by Chris Hostetter <ho...@fucit.org>.

: I am thinking to index these companies name in solr since all the functionality already there?
: 
: Do we have support for spark?

https://github.com/LucidWorks/spark-solr


Also of possible interest...

http://lucidworks.com/blog/solr-yarn/
https://github.com/LucidWorks/yarn-proto
https://issues.apache.org/jira/browse/SOLR-6743


-Hoss
http://www.lucidworks.com/