[jira] [Resolved] (IMPALA-2281) Use a better hash function than FNV for exchanges

["Tianyi Wang (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IMPALA-2281?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Tianyi Wang resolved IMPALA-2281.
---------------------------------
       Resolution: Fixed
    Fix Version/s: Impala 2.11.0

IMPALA-2281: Replace FNV with FastHash in exchange nodes

FNV is not a good enough hash function. This patch introduces FastHash
into the codebase and uses it in exchange nodes.

Testing: Two test cases involving arbitrary ordering are changed.
Single node performance benchmark shows no performance difference.

Change-Id: I778317d982dcdb94173a369a65b39f32b4f7ded2
Reviewed-on: http://gerrit.cloudera.org:8080/8417
Reviewed-by: Jim Apple <jb...@apache.org>
Tested-by: Impala Public Jenkins

> Use a better hash function than FNV for exchanges
> -------------------------------------------------
>
>                 Key: IMPALA-2281
>                 URL: https://issues.apache.org/jira/browse/IMPALA-2281
>             Project: IMPALA
>          Issue Type: Improvement
>          Components: Backend
>    Affects Versions: Impala 2.2
>            Reporter: Tim Armstrong
>            Assignee: Tianyi Wang
>            Priority: Minor
>             Fix For: Impala 2.11.0
>
>         Attachments: performance_result.txt
>
>
> The FNV hash function is used in Impala to hash data for exchanges and there are various utility functions for computing FNV hashes of different data types. There are clearly superior alternatives. For example, Murmur2 is already in the codebase, is faster to compute, and has better distribution properties.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)