You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@carbondata.apache.org by "kumar vishal (JIRA)" <ji...@apache.org> on 2018/11/21 07:04:00 UTC

[jira] [Created] (CARBONDATA-3113) Fixed Local Dictionary Query Performance and Added reusable buffer for direct flow

kumar vishal created CARBONDATA-3113:
----------------------------------------

             Summary: Fixed Local Dictionary Query Performance  and Added reusable buffer for direct flow
                 Key: CARBONDATA-3113
                 URL: https://issues.apache.org/jira/browse/CARBONDATA-3113
             Project: CarbonData
          Issue Type: Improvement
            Reporter: kumar vishal
            Assignee: kumar vishal


1. Added reusable buffer for direct flow
In query for each page each column it is creating a byte array, when number of columns are high it is causing lots of minor gc and degrading query performance, as each page is getting uncompressed one by one we can use same buffer for all the columns and based on requested size it will resize.

2. Fixed Local Dictionary performance issue.
Reverted back #2895 and fixed NPE issue by setting null for local dictionary to vector In safe and Unsafe VariableLengthDataChunkStore



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)