You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@carbondata.apache.org by "Ravindra Pesala (JIRA)" <ji...@apache.org> on 2018/10/16 07:09:00 UTC

[jira] [Created] (CARBONDATA-3010) Performance improvements for Fileformat and Presto

Ravindra Pesala created CARBONDATA-3010:
-------------------------------------------

             Summary: Performance improvements for Fileformat and Presto
                 Key: CARBONDATA-3010
                 URL: https://issues.apache.org/jira/browse/CARBONDATA-3010
             Project: CarbonData
          Issue Type: Bug
            Reporter: Ravindra Pesala


When querying data using Spark or Presto while filling the vector, carbondata is not well optimized. The major issues are as follows.
 # CarbonData has long method stack for reading and filling out the data to vector.
 # Many conditions and checks before filling out the data to vector.
 # Maintaining intermediate copies of data leads to more CPU utilization.
 # Filtering of data twice when using spark's fileformat vector flow and presto flow



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