[
https://issues.apache.org/jira/browse/CARBONDATA-2402?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16521772#comment-16521772 ]
xuchuanyin commented on CARBONDATA-2402:
----------------------------------------
It may affect backwards compatibility since pervious encoding use 8 bytes to encode value. Won't fix it now.
> Optimize allocated buffer size while converting object to byte array
> --------------------------------------------------------------------
>
> Key: CARBONDATA-2402
> URL:
https://issues.apache.org/jira/browse/CARBONDATA-2402> Project: CarbonData
> Issue Type: Improvement
> Reporter: xuchuanyin
> Assignee: xuchuanyin
> Priority: Major
> Time Spent: 50m
> Remaining Estimate: 0h
>
> Currently in carbondata, while converting object to byte array, it use 8 byte space for int and short datatype, which cause memory wasted.
> Code can be found in `CarbonUtil.{{getValueAsBytes` and `DataTypeUtil.getMeasureObjectFromDataType`}}
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)