drill-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Aman Sinha (JIRA)" <j...@apache.org>
Subject [jira] [Created] (DRILL-4891) Use field width when available for computing estimated memory costing
Date Wed, 14 Sep 2016 17:56:20 GMT
Aman Sinha created DRILL-4891:
---------------------------------

             Summary: Use field width when available for computing estimated memory costing
                 Key: DRILL-4891
                 URL: https://issues.apache.org/jira/browse/DRILL-4891
             Project: Apache Drill
          Issue Type: Improvement
          Components: Query Planning & Optimization
    Affects Versions: 1.8.0
            Reporter: Aman Sinha


Drill currently uses an average field width (in addition to other parameters) for computing
estimated memory cost for hash and sort based operations.  The assumption is that the field
is of ANY type.  However, when a view with appropriate CAST is used, the relevant data type
is available and we should leverage the correct field width for the costing calculations.




--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message