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

Posted to dev@parquet.apache.org by "Uwe L. Korn (JIRA)" <ji...@apache.org> on 2017/01/12 08:15:45 UTC

[jira] [Resolved] (PARQUET-817) [C++] Perform Parquet reads and Arrow deserialization concurrently

     [ https://issues.apache.org/jira/browse/PARQUET-817?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Uwe L. Korn resolved PARQUET-817.
---------------------------------

Closing this as directly reading into the Arrow layout is much faster.

We should still think of doing parallel processing as Parquet decoding is still CPU bound (independent of the target memory layout).

> [C++] Perform Parquet reads and Arrow deserialization concurrently
> ------------------------------------------------------------------
>
>                 Key: PARQUET-817
>                 URL: https://issues.apache.org/jira/browse/PARQUET-817
>             Project: Parquet
>          Issue Type: New Feature
>          Components: parquet-cpp
>            Reporter: Wes McKinney
>            Assignee: Uwe L. Korn
>
> The idea with this issue is to make batch requests from parquet-cpp happen asynchronously in a background thread, so we aren't pausing decoding while converting to Arrow representation



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)