You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@trafodion.apache.org by "ASF GitHub Bot (JIRA)" <ji...@apache.org> on 2016/03/02 23:56:18 UTC
[jira] [Commented] (TRAFODION-1867) UPDATE STATS forces serial
plans for non-partitioned tables
[ https://issues.apache.org/jira/browse/TRAFODION-1867?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15176654#comment-15176654 ]
ASF GitHub Bot commented on TRAFODION-1867:
-------------------------------------------
GitHub user DaveBirdsall opened a pull request:
https://github.com/apache/incubator-trafodion/pull/357
[TRAFODION-1867] Allow parallel plans for non-partitioned tables in U…
…PD STATS.
The problem is for large non-partitioned tables, UPDATE STATS will do a series of SELECTs to compute statistics for sets of columns. The plans for these SELECTs are presently forced to be serial for non-partitioned tables as a workaround for a long-ago compiler bug which seems to have been fixed. It's time to remove this workaround and thereby speed up UPDATE STATS.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/DaveBirdsall/incubator-trafodion SerialUpdateStatsBug
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-trafodion/pull/357.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #357
----
commit 07bf4163bf3c5b8bc863332886464e69f44cea6b
Author: Dave Birdsall <db...@apache.org>
Date: 2016-03-02T22:52:19Z
[TRAFODION-1867] Allow parallel plans for non-partitioned tables in UPD STATS
----
> UPDATE STATS forces serial plans for non-partitioned tables
> -----------------------------------------------------------
>
> Key: TRAFODION-1867
> URL: https://issues.apache.org/jira/browse/TRAFODION-1867
> Project: Apache Trafodion
> Issue Type: Bug
> Components: sql-cmu
> Affects Versions: 2.0-incubating, 1.3-incubating
> Environment: All
> Reporter: David Wayne Birdsall
> Assignee: David Wayne Birdsall
> Priority: Minor
>
> If one does an UPDATE STATISTICS for a non-partitioned table that is too large for internal sort, and does so without a SAMPLE clause, then UPDATE STATISTICS will perform SELECTs a few columns at a time, generating histograms on each. The plans are forced to be serial, even though the grouping and sorting phase could be done in parallel. This was evidently done as a workaround to a long-past compiler bug that seems to be fixed now. The serial plans slow down UPDATE STATISTICS, so it seems like it is time to remove the workaround code and let these plans be parallel.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)