Details
-
Sub-task
-
Status: Open
-
Minor
-
Resolution: Unresolved
-
None
-
None
-
None
-
None
Description
Thinking about this in the MapReduce world, it may also be worthwhile to run Embedded Solr rather than a separate SolrServer. I'm thinking this might map into jobs that can be split amongst N machines, each slicing up part of, say, the indexing process in which case it might be easier to use EmbeddedSolr. My naive thought is that these separate N indexes could be combined on a per-shard basis using MergeIndexes in the reduce step.
WARNING: I'm not at all sure how this will work, or whether it's desirable. This is more a placeholder to make sure we examine this possibility. See: http://wiki.apache.org/solr/EmbeddedSolr