You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hive.apache.org by Andrew Mains <an...@kontagent.com> on 2014/08/20 22:02:05 UTC

Re: hbase-handler predicate pushdown with multiple scans

If anyone needs it in the future, I've submitted a patch for this 
feature here: https://issues.apache.org/jira/browse/HIVE-7805

On 7/21/14, 5:26 PM, Andrew Mains wrote:
> Hi all,
>
> We have a table in hive/HBase with a composite row key, the first 
> field of which is a "bucket". Since the bucket is based on a hash, 
> every query we have on our data needs to search through each bucket, 
> and then to apply start and stop row filters within each bucket. The 
> most efficient way to do this in HBase would be to specify multiple 
> scans (one for each bucket), and apply the appropriate start and end 
> row key to each (using for instance 
> https://hbase.apache.org/apidocs/org/apache/hadoop/hbase/mapreduce/MultiTableInputFormat.html). 
> Has anybody used this sort of functionality/do people think this would 
> be useful? If there's nothing extant, and people would find this 
> helpful, I'd be more than happy to raise a JIRA and write a patch for 
> hbase-handler (I already have one in progress).
>
> Thanks!
>
> Andrew