You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Jason Dere (JIRA)" <ji...@apache.org> on 2018/04/02 22:21:00 UTC
[jira] [Updated] (HIVE-16282) Semijoin: Disable slow-start for the
bloom filter aggregate task
[ https://issues.apache.org/jira/browse/HIVE-16282?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jason Dere updated HIVE-16282:
------------------------------
Fix Version/s: (was: 2.3.0)
3.0.0
> Semijoin: Disable slow-start for the bloom filter aggregate task
> ----------------------------------------------------------------
>
> Key: HIVE-16282
> URL: https://issues.apache.org/jira/browse/HIVE-16282
> Project: Hive
> Issue Type: Bug
> Affects Versions: 2.2.0
> Reporter: Gopal V
> Assignee: Deepak Jaiswal
> Priority: Major
> Fix For: 3.0.0
>
> Attachments: HIVE-16282.1.patch, HIVE-16282.2.patch, HIVE-16282.3.patch, HIVE-16282.4.patch, HIVE-16282.5.patch, extended plan.rtf
>
>
> The slow-start of the bloom filter vertex is a scheduling problem which causes more pre-emption than is useful.
> When the bloom filters are arranged as follows
> Map 1(10 tasks)->Reducer 2(1 task)->Map 3(100 tasks)
> Map 3 and Map 1 are immediately active since Reducer 2 -> Map 3 is a broadcast edge.
> Once 3 tasks in Map 1 finish, the engine kills one active task from Map 3 to make room for Reducer 2.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)