[
https://issues.apache.org/jira/browse/SOLR-2604?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Doug Mittendorf updated SOLR-2604:
----------------------------------
Description:
Join support was recently added to Solr trunk (4.0) via SOLR-2272. We have a
need for this functionality, but don't want to upgrade to trunk, so I spent
some time back-porting the feature to work with the 3.x APIs. Not all of the
optimizations that Yonik implemented have been ported since the 3.x APIs don't
allow the same low-level access to the index, but the basic functionality is in
place and seems to work well when joining between an index with 1.2M docs and
another with 300k docs.
I still have yet to back-port the unit tests as those are heavily dependent on
other 4.0 code, but wanted to go ahead and post the initial code in case others
can make use of it.
was:
Join support was recently added to Solr trunk (4.0). We have a need for this
functionality, but don't want to upgrade to trunk, so I spent some time
back-porting the feature to work with the 3.x APIs. Not all of the
optimizations that Yonik implemented have been ported since the 3.x APIs don't
allow the same low-level access to the index, but the basic functionality is in
place and seems to work well when joining between an index with 1.2M docs and
another with 300k docs.
I still have yet to back-port the unit tests as those are heavily dependent on
other 4.0 code, but wanted to go ahead and post the initial code in case others
can make use of it.
> Backport Join support to Solr 3.x
> ---------------------------------
>
> Key: SOLR-2604
> URL: https://issues.apache.org/jira/browse/SOLR-2604
> Project: Solr
> Issue Type: New Feature
> Components: search
> Reporter: Doug Mittendorf
> Priority: Minor
> Fix For: 3.3
>
> Attachments: SOLR-2604.patch
>
>
> Join support was recently added to Solr trunk (4.0) via SOLR-2272. We have a
> need for this functionality, but don't want to upgrade to trunk, so I spent
> some time back-porting the feature to work with the 3.x APIs. Not all of the
> optimizations that Yonik implemented have been ported since the 3.x APIs
> don't allow the same low-level access to the index, but the basic
> functionality is in place and seems to work well when joining between an
> index with 1.2M docs and another with 300k docs.
> I still have yet to back-port the unit tests as those are heavily dependent
> on other 4.0 code, but wanted to go ahead and post the initial code in case
> others can make use of it.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
