[
https://issues.apache.org/jira/browse/BLUR-75?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13635855#comment-13635855
]
Tim Williams commented on BLUR-75:
----------------------------------
Can we move this to 0.2 to discuss? In my mind, such a fundamental [breaking]
change to the status quo should happen in 0.2... where we already have new
ground being broken. That said, it's an area that i already have to maintain a
patch locally for a custom row-summary record so i could just change my patch
to keep up my current behavior...
> Change the Thrift API - Change BlurResults to NOT include Rows or Records
> -------------------------------------------------------------------------
>
> Key: BLUR-75
> URL: https://issues.apache.org/jira/browse/BLUR-75
> Project: Apache Blur
> Issue Type: Bug
> Affects Versions: 0.1.5
> Reporter: Aaron McCurry
> Fix For: 0.1.5
>
>
> I would like to remove the ability to fetch data during the query call.
> Instead I would like to result rowids (when superquery is turned on) and
> rowids + recordids when (when superquery is turned off).
> The biggest reason for this change would be because the data being with the
> result can be so large that it runs the controller(s) out of memory. This
> happens because in the blurquery object the user can ask for N number of rows
> (or records is super is off) without knowing the size of the row or the size
> of all of the rows in the result.
> By not fetching the results in the query call, this will force a second call
> to fetch each row (multiple calls). This makes each request smaller and
> therefore the likely hood of running the controller out of memory is greatly
> reduced.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
