You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@parquet.apache.org by "ASF GitHub Bot (JIRA)" <ji...@apache.org> on 2019/02/27 08:16:00 UTC
[jira] [Commented] (PARQUET-1542) Merge multiple I/O to one time
I/O when read footer
[ https://issues.apache.org/jira/browse/PARQUET-1542?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16779022#comment-16779022 ]
ASF GitHub Bot commented on PARQUET-1542:
-----------------------------------------
wangshisan commented on pull request #624: [PARQUET-1542] Merge multiple I/O to one time I/O in method readFooter
URL: https://github.com/apache/parquet-mr/pull/624
Currently in method readFooter of class ParquetFileReader, it will invoke method converter.readParquetMetadata. To decode file metadata, there may be quite a number of times of I/O processing underlying, which may heavily slow down parquet read performance.
A better way is to read all the footer bytes in a single I/O, since we have already got footer length before, and pass down all the bytes to underlying.
And we have a test in our environment, this change may bring 50 percent improvement for some cases.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
> Merge multiple I/O to one time I/O when read footer
> ---------------------------------------------------
>
> Key: PARQUET-1542
> URL: https://issues.apache.org/jira/browse/PARQUET-1542
> Project: Parquet
> Issue Type: Improvement
> Components: parquet-mr
> Reporter: Wang, Gang
> Priority: Major
> Labels: pull-request-available
>
> Currently in method readFooter of class ParquetFileReader, it will invoke method *converter.readParquetMetadata.* To decode file metadata, there may be quite a number of times of I/O processing underlying, which may heavily slow down parquet read performance.
> A better way is to read all the footer bytes in a single I/O, since we have already got footer length before, and pass down all the bytes to underlying.
>
> And we have a test in our environment, this change may bring 50 percent improvement for some cases.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)