jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Edward Winston <ewins...@periscopeholdings.com>
Subject Inconsistent searching between cluster nodes
Date Mon, 03 Mar 2014 19:12:57 GMT
We have a web application with Jackrabbit integrated and we are having some
problems with full text searching.  Using the luke tool to inspect the
lucene indexes on the 2 nodes I can see that the data is in both, but when
I execute a query the results show on one node but not the other.

Here is a sample query :

select file.*, systemdoc.*, doctype.* 
from [b:SystemDocument] as systemdoc 
  inner join [b:DocumentType] as doctype
                on isdescendantnode(systemdoc, doctype) 
  inner join [nt:file] as file on isdescendantnode(file, systemdoc)
  inner join [nt:resource] as resource on ischildnode(resource, file)
 where (isDescendantNode(file, [/b/attachment/P/ADDJK])) 
   and contains(resource.*, 'Search Text')  
   and [file].[b:searchable] = true 
order by [systemdoc].[b:displayDocId], [file].[bso:fileName]

As stated, the query works on one node but not the other.  Interestingly,
when a file is added to the 2nd node, it still isn't searchable from the
second node but it is found with the same query on the first node.

We have observed that the generated indexes are slightly different between
the two nodes, but again the luke tool can find the text we are searching

Our clustering configuration uses file filesystem to store the data.  It's
a shared filesystem across the the 2 nodes.  If there is any other
information that is needed please let me know.

Edward Winston
"Trust me.  I know what I'm doing."
		-- Sledge Hammer

View raw message