drill-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Julian Hyde (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (DRILL-5293) Poor performance of Hash Table due to same hash value as distribution below
Date Thu, 23 Feb 2017 00:03:44 GMT

    [ https://issues.apache.org/jira/browse/DRILL-5293?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15879511#comment-15879511

Julian Hyde commented on DRILL-5293:

Or, let each hash function have a "generation", basically a seed, the same way that you would
if you are doing multiple hash partitioning runs. Each operator uses its operator ID as the
generation. That way you are only hashing once (and therefore you don't have to worry about
so much about the computational expense of the hash function).

> Poor performance of Hash Table due to same hash value as distribution below
> ---------------------------------------------------------------------------
>                 Key: DRILL-5293
>                 URL: https://issues.apache.org/jira/browse/DRILL-5293
>             Project: Apache Drill
>          Issue Type: Bug
>          Components: Execution - Codegen
>    Affects Versions: 1.8.0
>            Reporter: Boaz Ben-Zvi
>            Assignee: Boaz Ben-Zvi
> The computation of the hash value is basically the same whether for the Hash Table (used
by Hash Agg, and Hash Join), or for distribution of rows at the exchange. As a result, a specific
Hash Table (in a parallel minor fragment) gets only rows "filtered out" by the partition below
("upstream"), so the pattern of this filtering leads to a non uniform usage of the hash buckets
in the table.
>   Here is a simplified example: An exchange partitions into TWO (minor fragments), each
running a Hash Agg. So the partition sends rows of EVEN hash values to the first, and rows
of ODD hash values to the second. Now the first recomputes the _same_ hash value for its Hash
table -- and only the even buckets get used !!  (Or with a partition into EIGHT -- possibly
only one eighth of the buckets would be used !! ) 
>    This would lead to longer hash chains and thus a _poor performance_ !
> A possible solution -- add a distribution function distFunc (only for partitioning) that
takes the hash value and "scrambles" it so that the entropy in all the bits effects the low
bits of the output. This function should be applied (in HashPrelUtil) over the generated code
that produces the hash value, like:
>    distFunc( hash32(field1, hash32(field2, hash32(field3, 0))) );
> Tested with a huge hash aggregate (64 M rows) and a parallelism of 8 ( planner.width.max_per_node
= 8 ); minor fragments 0 and 4 used only 1/8 of their buckets, the others used 1/4 of their
buckets.  Maybe the reason for this variance is that distribution is using "hash32AsDouble"
and hash agg is using "hash32".  

This message was sent by Atlassian JIRA

View raw message