[
https://issues.apache.org/jira/browse/BLUR-139?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13686221#comment-13686221
]
Aaron McCurry commented on BLUR-139:
------------------------------------
The only way this can work is if the rowid is required in the query, which will
require us to parse the query at the controller. Currently the controllers
don't do anything with the queries but pass them on to the shard servers.
Perhaps a less invasive solution would be to put a check on the shards
themselves that will exit the query immediately if they don't have the rowid
that is required.
> Make querying by rowid less taxing on shard servers
> ---------------------------------------------------
>
> Key: BLUR-139
> URL: https://issues.apache.org/jira/browse/BLUR-139
> Project: Apache Blur
> Issue Type: Improvement
> Components: Blur
> Affects Versions: 0.2.0, 0.3.0
> Reporter: Jeff Borst
> Fix For: 0.2.0, 0.3.0
>
>
> Please add a mechanism that makes querying by rowid less taxing on the shard
> servers. Rowid is not something that I want to put in block cache, but it is
> a field that is convenient to use in queries.
> Currently in queries, rowid acts like any other field, so all shard servers
> that do not have the rowid in question have to parse the query and run the
> search even though they are guaranteed to not produce any results.
> It would be nice if this behavior could be short circuited when rowid is a
> required query element and is a full string (doesn't include wildcards or
> ranges of any sort).
--
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
