hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Namit Jain (JIRA)" <>
Subject [jira] [Commented] (HIVE-3652) Join optimization for star schema
Date Mon, 05 Nov 2012 06:01:15 GMT


Namit Jain commented on HIVE-3652:

I was thinking more from the point of the current implementation.
A backup task is per join operation currently.
Thinking more about it, we can have a backup task (which can be a tree of tasks).

It would be very difficult to fit the following in the current architecture. 
There are 10 dimension tables, 9 of them fit into memory and one of them dont.
Perform a map-only join for the first 9, and then a regular backup join for the last one.
I am not sure, if we want to optimize that.
> Join optimization for star schema
> ---------------------------------
>                 Key: HIVE-3652
>                 URL:
>             Project: Hive
>          Issue Type: Improvement
>          Components: Query Processor
>            Reporter: Amareshwari Sriramadasu
>            Assignee: Amareshwari Sriramadasu
> Currently, if we join one fact table with multiple dimension tables, it results in multiple
mapreduce jobs for each join with dimension table, because join would be on different keys
for each dimension. 
> Usually all the dimension tables will be small and can fit into memory and so map-side
join can used to join with fact table.
> In this issue I want to look at optimizing such query to generate single mapreduce job
sothat mapper loads dimension tables into memory and joins with fact table on different keys
as well.

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:

View raw message