[
https://issues.apache.org/jira/browse/TRAFODION-3096?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16511480#comment-16511480
]
ASF GitHub Bot commented on TRAFODION-3096:
-------------------------------------------
Github user asfgit closed the pull request at:
https://github.com/apache/trafodion/pull/1588
> clip the varchar type data from mxosrvr to jdbct4
> --------------------------------------------------
>
> Key: TRAFODION-3096
> URL: https://issues.apache.org/jira/browse/TRAFODION-3096
> Project: Apache Trafodion
> Issue Type: New Feature
> Components: client-jdbc-t4, connectivity-mxosrvr
> Reporter: LiZe
> Priority: Major
>
> When fetching varchar type data from SQL engine, mxosrvr will apply for
> enough memory to ensure that the data can be stored, regardless of the actual
> varchar length.Then send such a large piece of data to jdbct 4.This results
> in unnecessary network overhead and JDBC memory usage.
> So I added a function to clip the resulting data. The data to be sent to
> JDBCT4 in mxosrvr is trimmed to reduce network overhead and memory
> consumption.This function cuts out the redundant parts of varchar type data,
> and is very useful for the case where varchar is very long but actually
> occupies very little.
> I added a switch to jdbct4 code to turn this function on and off according to
> the situation.
> I tested my code with phx
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
