Reply
New Contributor
Posts: 4
Registered: ‎05-02-2016
Accepted Solution

Not able to index filename using Cloudera Solr MarR (file content is getting stored and searchable)

Not able to index filename using Cloudera Solr MarR (file content is getting stored and searchable):

 

1)      I am Using MapReduceIndexerTool to index various types of files (doc, pdf, xls etc)

${SOLR_HOME}/contrib/mr/search-mr-*-job.jar org.apache.solr.hadoop.MapReduceIndexerTool

 

2)      I am using morphline solrCell command chaining

fmap : { content : text, content-type : content_type }

 

3)      In schema.xml I have below entry

<field name="resourcename" type="text_general" indexed="true" stored="true"/>

<copyField source="resourcename" dest="text"/>

 

4)      If needed, I will provide the solrconfig.xml details

 

5)      I have below code

              IndexReader reader = DirectoryReader.open(rdir);

              Document doc = reader.document(0);

       System.out.println("Fields: " + doc.getFields());

 

What I observed is : the filename/resourcename has not been indexed.

I only see these fields:

content_type:application/pdf:

id:c385c455-5c1a-4284-937b-e88003fa3438#0:

author:Neogi, Anindya:

author_s:Neogi, Anindya:

last_modified:1461302622000:

_version_:1533216476522086400:

Highlighted
New Contributor
Posts: 4
Registered: ‎05-02-2016

Re: Not able to index filename using Cloudera Solr MarR (file content is getting stored and searchab

To get full hdfs filepath and filename into index, just add the following to schema.xml, then create collection and then index using mapR. Nothing needs to be specified in morphline config file.

   <field name="file_path" type="string" indexed="true" stored="true" />

   <field name="file_name" type="string" indexed="true" stored="true" />