[
https://issues.apache.org/jira/browse/IMPALA-9043?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Quanlong Huang updated IMPALA-9043:
-----------------------------------
Description:
The existing ORC lib provides limited interfaces for optimizations. E.g. no
interfaces for predicate pushdown.
We may need to implement our native ORC reader for thorough performance
improvement. There are some common logics with Parquet reader that we can
re-use, e.g. column encoding/decoding.
There is a related discussion in the email-list when we first supported reading
ORC:
[https://lists.apache.org/thread.html/bc692b9fb526b24499ca70e3f3f4fc258e1ec9edbc7bfe55c0a2b602%40%3Cdev.impala.apache.org%3E]
was:
The existing ORC lib provides limited interfaces for optimizations. E.g. no
interfaces for predicate pushdown.
We may need to implement our native ORC reader for thorough performance
improvement. There are some common logics with Parquet reader that we can
re-use, e.g. column encoding/decoding.
> Consider replacing the ORC lib with a native ORC reader
> -------------------------------------------------------
>
> Key: IMPALA-9043
> URL: https://issues.apache.org/jira/browse/IMPALA-9043
> Project: IMPALA
> Issue Type: New Feature
> Reporter: Quanlong Huang
> Priority: Major
>
> The existing ORC lib provides limited interfaces for optimizations. E.g. no
> interfaces for predicate pushdown.
> We may need to implement our native ORC reader for thorough performance
> improvement. There are some common logics with Parquet reader that we can
> re-use, e.g. column encoding/decoding.
> There is a related discussion in the email-list when we first supported
> reading ORC:
> [https://lists.apache.org/thread.html/bc692b9fb526b24499ca70e3f3f4fc258e1ec9edbc7bfe55c0a2b602%40%3Cdev.impala.apache.org%3E]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
